diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100755--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,30548 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.28736,
+  "global_step": 449000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 16.4528,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 11.3853,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5e-05,
+      "loss": 8.4021,
+      "step": 300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 7.2919,
+      "step": 400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5e-05,
+      "loss": 6.1762,
+      "step": 500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-05,
+      "loss": 5.2568,
+      "step": 600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5e-05,
+      "loss": 4.796,
+      "step": 700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4e-05,
+      "loss": 4.6566,
+      "step": 800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.5e-05,
+      "loss": 4.482,
+      "step": 900
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 4.2754,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 3.176161766052246,
+      "eval_runtime": 174.8066,
+      "eval_samples_per_second": 57.206,
+      "eval_steps_per_second": 3.575,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999679795068844e-05,
+      "loss": 4.1532,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999359590137688e-05,
+      "loss": 4.045,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999039385206533e-05,
+      "loss": 3.9962,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998719180275376e-05,
+      "loss": 3.9307,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998398975344221e-05,
+      "loss": 3.901,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998078770413064e-05,
+      "loss": 3.8329,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997758565481909e-05,
+      "loss": 3.8621,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997438360550753e-05,
+      "loss": 3.782,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997118155619597e-05,
+      "loss": 3.7489,
+      "step": 1900
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996797950688441e-05,
+      "loss": 3.7324,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 3.0246834754943848,
+      "eval_runtime": 174.5444,
+      "eval_samples_per_second": 57.292,
+      "eval_steps_per_second": 3.581,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9964777457572846e-05,
+      "loss": 3.6862,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996157540826129e-05,
+      "loss": 3.6695,
+      "step": 2200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9958373358949726e-05,
+      "loss": 3.644,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.995517130963817e-05,
+      "loss": 3.6352,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.995196926032661e-05,
+      "loss": 3.6435,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994876721101505e-05,
+      "loss": 3.5992,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994556516170349e-05,
+      "loss": 3.6032,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994236311239193e-05,
+      "loss": 3.5885,
+      "step": 2800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993916106308038e-05,
+      "loss": 3.5502,
+      "step": 2900
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993595901376881e-05,
+      "loss": 3.5497,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.9612457752227783,
+      "eval_runtime": 176.0693,
+      "eval_samples_per_second": 56.796,
+      "eval_steps_per_second": 3.55,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993275696445726e-05,
+      "loss": 3.5265,
+      "step": 3100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992955491514569e-05,
+      "loss": 3.5073,
+      "step": 3200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992635286583414e-05,
+      "loss": 3.5148,
+      "step": 3300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9923150816522576e-05,
+      "loss": 3.4933,
+      "step": 3400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9919948767211016e-05,
+      "loss": 3.5003,
+      "step": 3500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991674671789946e-05,
+      "loss": 3.4647,
+      "step": 3600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9913544668587896e-05,
+      "loss": 3.4672,
+      "step": 3700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991034261927634e-05,
+      "loss": 3.4315,
+      "step": 3800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9907140569964775e-05,
+      "loss": 3.4564,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990393852065322e-05,
+      "loss": 3.4795,
+      "step": 4000
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.9161300659179688,
+      "eval_runtime": 176.0774,
+      "eval_samples_per_second": 56.793,
+      "eval_steps_per_second": 3.55,
+      "step": 4000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990073647134166e-05,
+      "loss": 3.4406,
+      "step": 4100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98975344220301e-05,
+      "loss": 3.4532,
+      "step": 4200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989433237271854e-05,
+      "loss": 3.4261,
+      "step": 4300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989113032340698e-05,
+      "loss": 3.4327,
+      "step": 4400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988792827409543e-05,
+      "loss": 3.4158,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988472622478386e-05,
+      "loss": 3.3788,
+      "step": 4600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9881524175472307e-05,
+      "loss": 3.3911,
+      "step": 4700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987832212616074e-05,
+      "loss": 3.4317,
+      "step": 4800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9875120076849186e-05,
+      "loss": 3.3859,
+      "step": 4900
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9871918027537626e-05,
+      "loss": 3.3852,
+      "step": 5000
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.8852083683013916,
+      "eval_runtime": 175.4843,
+      "eval_samples_per_second": 56.985,
+      "eval_steps_per_second": 3.562,
+      "step": 5000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9868715978226066e-05,
+      "loss": 3.3619,
+      "step": 5100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986551392891451e-05,
+      "loss": 3.3791,
+      "step": 5200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9862311879602945e-05,
+      "loss": 3.3694,
+      "step": 5300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985910983029139e-05,
+      "loss": 3.3651,
+      "step": 5400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9855907780979825e-05,
+      "loss": 3.3703,
+      "step": 5500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985270573166827e-05,
+      "loss": 3.3411,
+      "step": 5600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984950368235671e-05,
+      "loss": 3.3514,
+      "step": 5700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984630163304515e-05,
+      "loss": 3.3375,
+      "step": 5800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984309958373359e-05,
+      "loss": 3.3522,
+      "step": 5900
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983989753442203e-05,
+      "loss": 3.3496,
+      "step": 6000
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.8661420345306396,
+      "eval_runtime": 175.3115,
+      "eval_samples_per_second": 57.041,
+      "eval_steps_per_second": 3.565,
+      "step": 6000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9836695485110477e-05,
+      "loss": 3.3567,
+      "step": 6100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983349343579891e-05,
+      "loss": 3.3431,
+      "step": 6200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9830291386487356e-05,
+      "loss": 3.3389,
+      "step": 6300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9827089337175796e-05,
+      "loss": 3.3292,
+      "step": 6400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823887287864235e-05,
+      "loss": 3.3314,
+      "step": 6500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9820685238552675e-05,
+      "loss": 3.319,
+      "step": 6600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9817483189241115e-05,
+      "loss": 3.3457,
+      "step": 6700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981428113992956e-05,
+      "loss": 3.3041,
+      "step": 6800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9811079090617994e-05,
+      "loss": 3.3096,
+      "step": 6900
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980787704130644e-05,
+      "loss": 3.2979,
+      "step": 7000
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.8510663509368896,
+      "eval_runtime": 176.6117,
+      "eval_samples_per_second": 56.621,
+      "eval_steps_per_second": 3.539,
+      "step": 7000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980467499199488e-05,
+      "loss": 3.3189,
+      "step": 7100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980147294268332e-05,
+      "loss": 3.3005,
+      "step": 7200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979827089337176e-05,
+      "loss": 3.2669,
+      "step": 7300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97950688440602e-05,
+      "loss": 3.2911,
+      "step": 7400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979186679474864e-05,
+      "loss": 3.3187,
+      "step": 7500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.978866474543708e-05,
+      "loss": 3.3035,
+      "step": 7600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9785462696125526e-05,
+      "loss": 3.2913,
+      "step": 7700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.978226064681396e-05,
+      "loss": 3.261,
+      "step": 7800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9779058597502405e-05,
+      "loss": 3.28,
+      "step": 7900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9775856548190845e-05,
+      "loss": 3.3035,
+      "step": 8000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.8327794075012207,
+      "eval_runtime": 175.8544,
+      "eval_samples_per_second": 56.865,
+      "eval_steps_per_second": 3.554,
+      "step": 8000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9772654498879285e-05,
+      "loss": 3.2916,
+      "step": 8100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9769452449567725e-05,
+      "loss": 3.2971,
+      "step": 8200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9766250400256164e-05,
+      "loss": 3.2899,
+      "step": 8300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.976304835094461e-05,
+      "loss": 3.2649,
+      "step": 8400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9759846301633044e-05,
+      "loss": 3.2695,
+      "step": 8500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.975664425232149e-05,
+      "loss": 3.2501,
+      "step": 8600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.975344220300993e-05,
+      "loss": 3.2668,
+      "step": 8700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.975024015369837e-05,
+      "loss": 3.2317,
+      "step": 8800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974703810438681e-05,
+      "loss": 3.2606,
+      "step": 8900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974383605507525e-05,
+      "loss": 3.264,
+      "step": 9000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.8224401473999023,
+      "eval_runtime": 177.2958,
+      "eval_samples_per_second": 56.403,
+      "eval_steps_per_second": 3.525,
+      "step": 9000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974063400576369e-05,
+      "loss": 3.2375,
+      "step": 9100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973743195645213e-05,
+      "loss": 3.2443,
+      "step": 9200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9734229907140575e-05,
+      "loss": 3.2454,
+      "step": 9300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731027857829015e-05,
+      "loss": 3.2352,
+      "step": 9400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9727825808517455e-05,
+      "loss": 3.2763,
+      "step": 9500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9724623759205894e-05,
+      "loss": 3.2415,
+      "step": 9600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9721421709894334e-05,
+      "loss": 3.2385,
+      "step": 9700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9718219660582774e-05,
+      "loss": 3.2458,
+      "step": 9800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9715017611271214e-05,
+      "loss": 3.267,
+      "step": 9900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971181556195966e-05,
+      "loss": 3.2282,
+      "step": 10000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.811612367630005,
+      "eval_runtime": 175.957,
+      "eval_samples_per_second": 56.832,
+      "eval_steps_per_second": 3.552,
+      "step": 10000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970861351264809e-05,
+      "loss": 3.2373,
+      "step": 10100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970541146333654e-05,
+      "loss": 3.2365,
+      "step": 10200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970220941402498e-05,
+      "loss": 3.2106,
+      "step": 10300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969900736471342e-05,
+      "loss": 3.2247,
+      "step": 10400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969580531540186e-05,
+      "loss": 3.2031,
+      "step": 10500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96926032660903e-05,
+      "loss": 3.2485,
+      "step": 10600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968940121677874e-05,
+      "loss": 3.2195,
+      "step": 10700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968619916746718e-05,
+      "loss": 3.2253,
+      "step": 10800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9682997118155625e-05,
+      "loss": 3.209,
+      "step": 10900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9679795068844064e-05,
+      "loss": 3.1864,
+      "step": 11000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7983367443084717,
+      "eval_runtime": 176.0989,
+      "eval_samples_per_second": 56.786,
+      "eval_steps_per_second": 3.549,
+      "step": 11000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9676593019532504e-05,
+      "loss": 3.2072,
+      "step": 11100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9673390970220944e-05,
+      "loss": 3.1995,
+      "step": 11200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9670188920909384e-05,
+      "loss": 3.1948,
+      "step": 11300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966698687159782e-05,
+      "loss": 3.1858,
+      "step": 11400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966378482228626e-05,
+      "loss": 3.1929,
+      "step": 11500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966058277297471e-05,
+      "loss": 3.17,
+      "step": 11600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965738072366315e-05,
+      "loss": 3.1996,
+      "step": 11700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965417867435159e-05,
+      "loss": 3.1755,
+      "step": 11800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965097662504003e-05,
+      "loss": 3.1663,
+      "step": 11900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964777457572847e-05,
+      "loss": 3.1999,
+      "step": 12000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7982311248779297,
+      "eval_runtime": 177.464,
+      "eval_samples_per_second": 56.349,
+      "eval_steps_per_second": 3.522,
+      "step": 12000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964457252641691e-05,
+      "loss": 3.1752,
+      "step": 12100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964137047710535e-05,
+      "loss": 3.1834,
+      "step": 12200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963816842779379e-05,
+      "loss": 3.1905,
+      "step": 12300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963496637848223e-05,
+      "loss": 3.1793,
+      "step": 12400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9631764329170674e-05,
+      "loss": 3.1939,
+      "step": 12500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9628562279859114e-05,
+      "loss": 3.1722,
+      "step": 12600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9625360230547553e-05,
+      "loss": 3.1581,
+      "step": 12700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962215818123599e-05,
+      "loss": 3.1796,
+      "step": 12800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961895613192443e-05,
+      "loss": 3.2063,
+      "step": 12900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961575408261287e-05,
+      "loss": 3.1999,
+      "step": 13000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7836506366729736,
+      "eval_runtime": 175.8615,
+      "eval_samples_per_second": 56.863,
+      "eval_steps_per_second": 3.554,
+      "step": 13000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961255203330131e-05,
+      "loss": 3.179,
+      "step": 13100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960934998398976e-05,
+      "loss": 3.1719,
+      "step": 13200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96061479346782e-05,
+      "loss": 3.181,
+      "step": 13300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960294588536664e-05,
+      "loss": 3.1698,
+      "step": 13400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959974383605508e-05,
+      "loss": 3.1541,
+      "step": 13500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959654178674352e-05,
+      "loss": 3.1493,
+      "step": 13600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959333973743196e-05,
+      "loss": 3.1774,
+      "step": 13700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95901376881204e-05,
+      "loss": 3.1627,
+      "step": 13800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958693563880884e-05,
+      "loss": 3.1446,
+      "step": 13900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9583733589497284e-05,
+      "loss": 3.1574,
+      "step": 14000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.776357412338257,
+      "eval_runtime": 176.1722,
+      "eval_samples_per_second": 56.763,
+      "eval_steps_per_second": 3.548,
+      "step": 14000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958053154018572e-05,
+      "loss": 3.1542,
+      "step": 14100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957732949087416e-05,
+      "loss": 3.1338,
+      "step": 14200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95741274415626e-05,
+      "loss": 3.1304,
+      "step": 14300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957092539225104e-05,
+      "loss": 3.1243,
+      "step": 14400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956772334293948e-05,
+      "loss": 3.1412,
+      "step": 14500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956452129362792e-05,
+      "loss": 3.174,
+      "step": 14600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956131924431636e-05,
+      "loss": 3.1614,
+      "step": 14700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955811719500481e-05,
+      "loss": 3.1496,
+      "step": 14800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955491514569325e-05,
+      "loss": 3.1417,
+      "step": 14900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955171309638169e-05,
+      "loss": 3.1302,
+      "step": 15000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7699687480926514,
+      "eval_runtime": 175.7202,
+      "eval_samples_per_second": 56.909,
+      "eval_steps_per_second": 3.557,
+      "step": 15000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954851104707013e-05,
+      "loss": 3.1279,
+      "step": 15100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954530899775857e-05,
+      "loss": 3.1472,
+      "step": 15200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954210694844701e-05,
+      "loss": 3.1461,
+      "step": 15300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890489913545e-05,
+      "loss": 3.1369,
+      "step": 15400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9535702849823886e-05,
+      "loss": 3.1469,
+      "step": 15500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953250080051233e-05,
+      "loss": 3.1522,
+      "step": 15600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952929875120077e-05,
+      "loss": 3.1273,
+      "step": 15700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952609670188921e-05,
+      "loss": 3.1445,
+      "step": 15800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952289465257765e-05,
+      "loss": 3.1391,
+      "step": 15900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951969260326609e-05,
+      "loss": 3.1212,
+      "step": 16000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7630386352539062,
+      "eval_runtime": 176.7557,
+      "eval_samples_per_second": 56.575,
+      "eval_steps_per_second": 3.536,
+      "step": 16000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951649055395453e-05,
+      "loss": 3.1535,
+      "step": 16100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951328850464297e-05,
+      "loss": 3.127,
+      "step": 16200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951008645533142e-05,
+      "loss": 3.1363,
+      "step": 16300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950688440601986e-05,
+      "loss": 3.1224,
+      "step": 16400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95036823567083e-05,
+      "loss": 3.1081,
+      "step": 16500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950048030739674e-05,
+      "loss": 3.1115,
+      "step": 16600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949727825808518e-05,
+      "loss": 3.1432,
+      "step": 16700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949407620877362e-05,
+      "loss": 3.1042,
+      "step": 16800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9490874159462056e-05,
+      "loss": 3.1345,
+      "step": 16900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94876721101505e-05,
+      "loss": 3.1021,
+      "step": 17000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.75553560256958,
+      "eval_runtime": 176.7294,
+      "eval_samples_per_second": 56.584,
+      "eval_steps_per_second": 3.536,
+      "step": 17000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9484470060838936e-05,
+      "loss": 3.0994,
+      "step": 17100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948126801152738e-05,
+      "loss": 3.1197,
+      "step": 17200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947806596221582e-05,
+      "loss": 3.1163,
+      "step": 17300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947486391290426e-05,
+      "loss": 3.1142,
+      "step": 17400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94716618635927e-05,
+      "loss": 3.115,
+      "step": 17500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946845981428114e-05,
+      "loss": 3.118,
+      "step": 17600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946525776496958e-05,
+      "loss": 3.1086,
+      "step": 17700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946205571565802e-05,
+      "loss": 3.0966,
+      "step": 17800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945885366634647e-05,
+      "loss": 3.1024,
+      "step": 17900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945565161703491e-05,
+      "loss": 3.1026,
+      "step": 18000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7484090328216553,
+      "eval_runtime": 175.5292,
+      "eval_samples_per_second": 56.971,
+      "eval_steps_per_second": 3.561,
+      "step": 18000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945244956772335e-05,
+      "loss": 3.1031,
+      "step": 18100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9449247518411787e-05,
+      "loss": 3.0998,
+      "step": 18200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446045469100226e-05,
+      "loss": 3.1153,
+      "step": 18300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9442843419788666e-05,
+      "loss": 3.1114,
+      "step": 18400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9439641370477106e-05,
+      "loss": 3.1124,
+      "step": 18500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943643932116555e-05,
+      "loss": 3.1088,
+      "step": 18600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9433237271853985e-05,
+      "loss": 3.0875,
+      "step": 18700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943003522254243e-05,
+      "loss": 3.1066,
+      "step": 18800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942683317323087e-05,
+      "loss": 3.0911,
+      "step": 18900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942363112391931e-05,
+      "loss": 3.1073,
+      "step": 19000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.745687484741211,
+      "eval_runtime": 178.9968,
+      "eval_samples_per_second": 55.867,
+      "eval_steps_per_second": 3.492,
+      "step": 19000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942042907460775e-05,
+      "loss": 3.0936,
+      "step": 19100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941722702529619e-05,
+      "loss": 3.0894,
+      "step": 19200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941402497598464e-05,
+      "loss": 3.0944,
+      "step": 19300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941082292667307e-05,
+      "loss": 3.065,
+      "step": 19400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940762087736152e-05,
+      "loss": 3.0982,
+      "step": 19500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404418828049956e-05,
+      "loss": 3.0787,
+      "step": 19600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9401216778738396e-05,
+      "loss": 3.0753,
+      "step": 19700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9398014729426836e-05,
+      "loss": 3.0793,
+      "step": 19800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9394812680115276e-05,
+      "loss": 3.0907,
+      "step": 19900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9391610630803715e-05,
+      "loss": 3.099,
+      "step": 20000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.741987466812134,
+      "eval_runtime": 177.0933,
+      "eval_samples_per_second": 56.467,
+      "eval_steps_per_second": 3.529,
+      "step": 20000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388408581492155e-05,
+      "loss": 3.0733,
+      "step": 20100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93852065321806e-05,
+      "loss": 3.0774,
+      "step": 20200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9382004482869035e-05,
+      "loss": 3.1003,
+      "step": 20300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937880243355748e-05,
+      "loss": 3.08,
+      "step": 20400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937560038424592e-05,
+      "loss": 3.0872,
+      "step": 20500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937239833493436e-05,
+      "loss": 3.0771,
+      "step": 20600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93691962856228e-05,
+      "loss": 3.0922,
+      "step": 20700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936599423631124e-05,
+      "loss": 3.0874,
+      "step": 20800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362792186999687e-05,
+      "loss": 3.0597,
+      "step": 20900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935959013768812e-05,
+      "loss": 3.065,
+      "step": 21000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7343788146972656,
+      "eval_runtime": 177.3457,
+      "eval_samples_per_second": 56.387,
+      "eval_steps_per_second": 3.524,
+      "step": 21000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9356388088376566e-05,
+      "loss": 3.0777,
+      "step": 21100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9353186039065006e-05,
+      "loss": 3.075,
+      "step": 21200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9349983989753446e-05,
+      "loss": 3.052,
+      "step": 21300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346781940441885e-05,
+      "loss": 3.0808,
+      "step": 21400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9343579891130325e-05,
+      "loss": 3.0748,
+      "step": 21500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934037784181877e-05,
+      "loss": 3.0511,
+      "step": 21600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337175792507204e-05,
+      "loss": 3.0571,
+      "step": 21700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933397374319565e-05,
+      "loss": 3.0703,
+      "step": 21800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9330771693884084e-05,
+      "loss": 3.0712,
+      "step": 21900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932756964457253e-05,
+      "loss": 3.069,
+      "step": 22000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7288901805877686,
+      "eval_runtime": 177.5959,
+      "eval_samples_per_second": 56.308,
+      "eval_steps_per_second": 3.519,
+      "step": 22000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932436759526097e-05,
+      "loss": 3.0585,
+      "step": 22100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932116554594941e-05,
+      "loss": 3.0851,
+      "step": 22200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931796349663785e-05,
+      "loss": 3.0438,
+      "step": 22300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931476144732629e-05,
+      "loss": 3.0897,
+      "step": 22400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9311559398014736e-05,
+      "loss": 3.0659,
+      "step": 22500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930835734870317e-05,
+      "loss": 3.0311,
+      "step": 22600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9305155299391615e-05,
+      "loss": 3.0685,
+      "step": 22700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9301953250080055e-05,
+      "loss": 3.0618,
+      "step": 22800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9298751200768495e-05,
+      "loss": 3.0552,
+      "step": 22900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295549151456935e-05,
+      "loss": 3.0592,
+      "step": 23000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 2.7260777950286865,
+      "eval_runtime": 179.666,
+      "eval_samples_per_second": 55.659,
+      "eval_steps_per_second": 3.479,
+      "step": 23000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9292347102145374e-05,
+      "loss": 3.0724,
+      "step": 23100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928914505283382e-05,
+      "loss": 3.0681,
+      "step": 23200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9285943003522254e-05,
+      "loss": 3.0484,
+      "step": 23300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92827409542107e-05,
+      "loss": 3.0626,
+      "step": 23400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.927953890489913e-05,
+      "loss": 3.0594,
+      "step": 23500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.927633685558758e-05,
+      "loss": 3.0619,
+      "step": 23600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.927313480627602e-05,
+      "loss": 3.0485,
+      "step": 23700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.926993275696446e-05,
+      "loss": 3.0581,
+      "step": 23800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9266730707652906e-05,
+      "loss": 3.0889,
+      "step": 23900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.926352865834134e-05,
+      "loss": 3.0676,
+      "step": 24000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.72514271736145,
+      "eval_runtime": 176.9986,
+      "eval_samples_per_second": 56.498,
+      "eval_steps_per_second": 3.531,
+      "step": 24000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9260326609029785e-05,
+      "loss": 3.045,
+      "step": 24100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.925712455971822e-05,
+      "loss": 3.0581,
+      "step": 24200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9253922510406665e-05,
+      "loss": 3.0509,
+      "step": 24300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9250720461095105e-05,
+      "loss": 3.0461,
+      "step": 24400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9247518411783544e-05,
+      "loss": 3.0504,
+      "step": 24500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9244316362471984e-05,
+      "loss": 3.0376,
+      "step": 24600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9241114313160424e-05,
+      "loss": 3.0529,
+      "step": 24700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923791226384887e-05,
+      "loss": 3.0611,
+      "step": 24800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92347102145373e-05,
+      "loss": 3.0552,
+      "step": 24900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923150816522575e-05,
+      "loss": 3.0424,
+      "step": 25000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.7178871631622314,
+      "eval_runtime": 177.5357,
+      "eval_samples_per_second": 56.327,
+      "eval_steps_per_second": 3.52,
+      "step": 25000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922830611591418e-05,
+      "loss": 3.0364,
+      "step": 25100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922510406660263e-05,
+      "loss": 3.05,
+      "step": 25200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922190201729107e-05,
+      "loss": 3.0679,
+      "step": 25300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921869996797951e-05,
+      "loss": 3.0362,
+      "step": 25400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9215497918667955e-05,
+      "loss": 3.028,
+      "step": 25500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921229586935639e-05,
+      "loss": 3.0405,
+      "step": 25600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9209093820044835e-05,
+      "loss": 3.046,
+      "step": 25700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920589177073327e-05,
+      "loss": 3.0299,
+      "step": 25800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9202689721421714e-05,
+      "loss": 3.0186,
+      "step": 25900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9199487672110154e-05,
+      "loss": 3.0578,
+      "step": 26000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.713019609451294,
+      "eval_runtime": 178.6143,
+      "eval_samples_per_second": 55.987,
+      "eval_steps_per_second": 3.499,
+      "step": 26000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9196285622798594e-05,
+      "loss": 3.0344,
+      "step": 26100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919308357348703e-05,
+      "loss": 3.0358,
+      "step": 26200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918988152417547e-05,
+      "loss": 3.0481,
+      "step": 26300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918667947486392e-05,
+      "loss": 3.0317,
+      "step": 26400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918347742555235e-05,
+      "loss": 3.0464,
+      "step": 26500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91802753762408e-05,
+      "loss": 3.0183,
+      "step": 26600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917707332692923e-05,
+      "loss": 3.0408,
+      "step": 26700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917387127761768e-05,
+      "loss": 3.0367,
+      "step": 26800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917066922830612e-05,
+      "loss": 3.015,
+      "step": 26900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916746717899456e-05,
+      "loss": 3.0268,
+      "step": 27000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.709913730621338,
+      "eval_runtime": 177.2666,
+      "eval_samples_per_second": 56.412,
+      "eval_steps_per_second": 3.526,
+      "step": 27000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9164265129683005e-05,
+      "loss": 3.035,
+      "step": 27100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916106308037144e-05,
+      "loss": 3.0458,
+      "step": 27200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9157861031059884e-05,
+      "loss": 3.034,
+      "step": 27300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915465898174832e-05,
+      "loss": 3.0237,
+      "step": 27400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9151456932436764e-05,
+      "loss": 3.0161,
+      "step": 27500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9148254883125197e-05,
+      "loss": 3.017,
+      "step": 27600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914505283381364e-05,
+      "loss": 3.0347,
+      "step": 27700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914185078450208e-05,
+      "loss": 3.0212,
+      "step": 27800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913864873519052e-05,
+      "loss": 3.0185,
+      "step": 27900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913544668587897e-05,
+      "loss": 3.0408,
+      "step": 28000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.7085154056549072,
+      "eval_runtime": 175.5048,
+      "eval_samples_per_second": 56.978,
+      "eval_steps_per_second": 3.561,
+      "step": 28000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91322446365674e-05,
+      "loss": 3.0364,
+      "step": 28100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912904258725585e-05,
+      "loss": 3.0225,
+      "step": 28200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912584053794428e-05,
+      "loss": 3.0096,
+      "step": 28300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912263848863273e-05,
+      "loss": 3.0086,
+      "step": 28400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911943643932117e-05,
+      "loss": 3.0182,
+      "step": 28500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911623439000961e-05,
+      "loss": 3.0092,
+      "step": 28600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9113032340698054e-05,
+      "loss": 3.0079,
+      "step": 28700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910983029138649e-05,
+      "loss": 2.9888,
+      "step": 28800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9106628242074933e-05,
+      "loss": 3.0265,
+      "step": 28900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103426192763366e-05,
+      "loss": 3.0221,
+      "step": 29000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.7025697231292725,
+      "eval_runtime": 177.1206,
+      "eval_samples_per_second": 56.459,
+      "eval_steps_per_second": 3.529,
+      "step": 29000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910022414345181e-05,
+      "loss": 3.0081,
+      "step": 29100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909702209414025e-05,
+      "loss": 3.029,
+      "step": 29200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909382004482869e-05,
+      "loss": 3.0081,
+      "step": 29300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909061799551713e-05,
+      "loss": 3.0051,
+      "step": 29400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908741594620557e-05,
+      "loss": 3.0204,
+      "step": 29500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908421389689402e-05,
+      "loss": 3.0034,
+      "step": 29600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908101184758245e-05,
+      "loss": 3.0299,
+      "step": 29700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90778097982709e-05,
+      "loss": 3.0136,
+      "step": 29800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907460774895933e-05,
+      "loss": 3.0318,
+      "step": 29900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907140569964778e-05,
+      "loss": 3.0013,
+      "step": 30000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.7001781463623047,
+      "eval_runtime": 178.1965,
+      "eval_samples_per_second": 56.118,
+      "eval_steps_per_second": 3.507,
+      "step": 30000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906820365033622e-05,
+      "loss": 3.0124,
+      "step": 30100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906500160102466e-05,
+      "loss": 3.0026,
+      "step": 30200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90617995517131e-05,
+      "loss": 3.0211,
+      "step": 30300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9058597502401536e-05,
+      "loss": 3.0049,
+      "step": 30400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905539545308998e-05,
+      "loss": 3.0135,
+      "step": 30500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052193403778416e-05,
+      "loss": 2.9869,
+      "step": 30600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904899135446686e-05,
+      "loss": 2.9998,
+      "step": 30700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90457893051553e-05,
+      "loss": 3.0038,
+      "step": 30800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904258725584374e-05,
+      "loss": 3.0082,
+      "step": 30900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903938520653218e-05,
+      "loss": 3.0048,
+      "step": 31000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.6981754302978516,
+      "eval_runtime": 176.8202,
+      "eval_samples_per_second": 56.555,
+      "eval_steps_per_second": 3.535,
+      "step": 31000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903618315722062e-05,
+      "loss": 3.0059,
+      "step": 31100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903298110790907e-05,
+      "loss": 3.0255,
+      "step": 31200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90297790585975e-05,
+      "loss": 2.9953,
+      "step": 31300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902657700928595e-05,
+      "loss": 2.9831,
+      "step": 31400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902337495997439e-05,
+      "loss": 2.9981,
+      "step": 31500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902017291066283e-05,
+      "loss": 2.9991,
+      "step": 31600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9016970861351266e-05,
+      "loss": 2.9946,
+      "step": 31700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9013768812039706e-05,
+      "loss": 3.0027,
+      "step": 31800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901056676272815e-05,
+      "loss": 2.971,
+      "step": 31900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9007364713416586e-05,
+      "loss": 3.0191,
+      "step": 32000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.6920597553253174,
+      "eval_runtime": 176.7041,
+      "eval_samples_per_second": 56.592,
+      "eval_steps_per_second": 3.537,
+      "step": 32000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900416266410503e-05,
+      "loss": 2.9977,
+      "step": 32100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9000960614793465e-05,
+      "loss": 3.0034,
+      "step": 32200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899775856548191e-05,
+      "loss": 2.9882,
+      "step": 32300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899455651617035e-05,
+      "loss": 2.9881,
+      "step": 32400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899135446685879e-05,
+      "loss": 2.9803,
+      "step": 32500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898815241754723e-05,
+      "loss": 2.9922,
+      "step": 32600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898495036823567e-05,
+      "loss": 2.9709,
+      "step": 32700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898174831892412e-05,
+      "loss": 2.9976,
+      "step": 32800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897854626961255e-05,
+      "loss": 2.9985,
+      "step": 32900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8975344220300997e-05,
+      "loss": 3.0019,
+      "step": 33000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.692898988723755,
+      "eval_runtime": 177.3804,
+      "eval_samples_per_second": 56.376,
+      "eval_steps_per_second": 3.524,
+      "step": 33000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8972142170989436e-05,
+      "loss": 2.9773,
+      "step": 33100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968940121677876e-05,
+      "loss": 2.9913,
+      "step": 33200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8965738072366316e-05,
+      "loss": 2.9822,
+      "step": 33300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8962536023054756e-05,
+      "loss": 3.0064,
+      "step": 33400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89593339737432e-05,
+      "loss": 2.989,
+      "step": 33500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8956131924431635e-05,
+      "loss": 2.9827,
+      "step": 33600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895292987512008e-05,
+      "loss": 3.0022,
+      "step": 33700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894972782580852e-05,
+      "loss": 2.9779,
+      "step": 33800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894652577649696e-05,
+      "loss": 2.9612,
+      "step": 33900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89433237271854e-05,
+      "loss": 2.9764,
+      "step": 34000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.688096761703491,
+      "eval_runtime": 177.6839,
+      "eval_samples_per_second": 56.28,
+      "eval_steps_per_second": 3.517,
+      "step": 34000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894012167787384e-05,
+      "loss": 2.9807,
+      "step": 34100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893691962856228e-05,
+      "loss": 3.0115,
+      "step": 34200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893371757925072e-05,
+      "loss": 2.9909,
+      "step": 34300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8930515529939166e-05,
+      "loss": 3.007,
+      "step": 34400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8927313480627606e-05,
+      "loss": 2.9952,
+      "step": 34500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8924111431316046e-05,
+      "loss": 2.9875,
+      "step": 34600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8920909382004486e-05,
+      "loss": 2.9829,
+      "step": 34700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8917707332692925e-05,
+      "loss": 2.9682,
+      "step": 34800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8914505283381365e-05,
+      "loss": 2.9665,
+      "step": 34900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8911303234069805e-05,
+      "loss": 2.9666,
+      "step": 35000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.6841659545898438,
+      "eval_runtime": 177.4963,
+      "eval_samples_per_second": 56.339,
+      "eval_steps_per_second": 3.521,
+      "step": 35000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890810118475825e-05,
+      "loss": 2.9709,
+      "step": 35100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8904899135446684e-05,
+      "loss": 2.9748,
+      "step": 35200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890169708613513e-05,
+      "loss": 2.9516,
+      "step": 35300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889849503682357e-05,
+      "loss": 2.9764,
+      "step": 35400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889529298751201e-05,
+      "loss": 2.9867,
+      "step": 35500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889209093820045e-05,
+      "loss": 3.0054,
+      "step": 35600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 2.9506,
+      "step": 35700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888568683957733e-05,
+      "loss": 2.976,
+      "step": 35800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888248479026577e-05,
+      "loss": 2.9867,
+      "step": 35900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8879282740954216e-05,
+      "loss": 2.9753,
+      "step": 36000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.68265438079834,
+      "eval_runtime": 175.295,
+      "eval_samples_per_second": 57.047,
+      "eval_steps_per_second": 3.565,
+      "step": 36000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876080691642656e-05,
+      "loss": 2.9837,
+      "step": 36100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8872878642331095e-05,
+      "loss": 2.9745,
+      "step": 36200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8869676593019535e-05,
+      "loss": 2.9655,
+      "step": 36300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8866474543707975e-05,
+      "loss": 2.979,
+      "step": 36400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8863272494396415e-05,
+      "loss": 2.982,
+      "step": 36500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8860070445084854e-05,
+      "loss": 2.9916,
+      "step": 36600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88568683957733e-05,
+      "loss": 2.9592,
+      "step": 36700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885366634646174e-05,
+      "loss": 2.961,
+      "step": 36800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885046429715018e-05,
+      "loss": 2.961,
+      "step": 36900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884726224783862e-05,
+      "loss": 2.9635,
+      "step": 37000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.6792824268341064,
+      "eval_runtime": 178.649,
+      "eval_samples_per_second": 55.976,
+      "eval_steps_per_second": 3.498,
+      "step": 37000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884406019852706e-05,
+      "loss": 2.9596,
+      "step": 37100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88408581492155e-05,
+      "loss": 2.9909,
+      "step": 37200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883765609990394e-05,
+      "loss": 2.9932,
+      "step": 37300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883445405059238e-05,
+      "loss": 2.9736,
+      "step": 37400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883125200128082e-05,
+      "loss": 2.974,
+      "step": 37500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8828049951969265e-05,
+      "loss": 2.9549,
+      "step": 37600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8824847902657705e-05,
+      "loss": 2.9616,
+      "step": 37700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8821645853346145e-05,
+      "loss": 2.9713,
+      "step": 37800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8818443804034584e-05,
+      "loss": 2.9662,
+      "step": 37900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8815241754723024e-05,
+      "loss": 2.978,
+      "step": 38000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.6746363639831543,
+      "eval_runtime": 177.2123,
+      "eval_samples_per_second": 56.429,
+      "eval_steps_per_second": 3.527,
+      "step": 38000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8812039705411464e-05,
+      "loss": 2.9669,
+      "step": 38100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8808837656099904e-05,
+      "loss": 2.9603,
+      "step": 38200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880563560678835e-05,
+      "loss": 2.9543,
+      "step": 38300
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880243355747679e-05,
+      "loss": 2.9636,
+      "step": 38400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879923150816523e-05,
+      "loss": 2.9742,
+      "step": 38500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879602945885367e-05,
+      "loss": 2.9462,
+      "step": 38600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879282740954211e-05,
+      "loss": 2.9654,
+      "step": 38700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878962536023055e-05,
+      "loss": 2.9638,
+      "step": 38800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878642331091899e-05,
+      "loss": 2.9718,
+      "step": 38900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878322126160743e-05,
+      "loss": 2.9599,
+      "step": 39000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 2.67553448677063,
+      "eval_runtime": 178.0811,
+      "eval_samples_per_second": 56.154,
+      "eval_steps_per_second": 3.51,
+      "step": 39000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8780019212295875e-05,
+      "loss": 2.9634,
+      "step": 39100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8776817162984315e-05,
+      "loss": 2.9718,
+      "step": 39200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8773615113672754e-05,
+      "loss": 2.9626,
+      "step": 39300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8770413064361194e-05,
+      "loss": 2.9759,
+      "step": 39400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8767211015049634e-05,
+      "loss": 2.9666,
+      "step": 39500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8764008965738074e-05,
+      "loss": 2.9583,
+      "step": 39600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.876080691642651e-05,
+      "loss": 2.9415,
+      "step": 39700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.875760486711495e-05,
+      "loss": 2.9487,
+      "step": 39800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.87544028178034e-05,
+      "loss": 2.9589,
+      "step": 39900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.875120076849184e-05,
+      "loss": 2.9576,
+      "step": 40000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6733338832855225,
+      "eval_runtime": 175.3918,
+      "eval_samples_per_second": 57.015,
+      "eval_steps_per_second": 3.563,
+      "step": 40000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.874799871918028e-05,
+      "loss": 2.9754,
+      "step": 40100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.874479666986872e-05,
+      "loss": 2.97,
+      "step": 40200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.874159462055716e-05,
+      "loss": 2.9666,
+      "step": 40300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.87383925712456e-05,
+      "loss": 2.9546,
+      "step": 40400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.873519052193404e-05,
+      "loss": 2.9265,
+      "step": 40500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.873198847262248e-05,
+      "loss": 2.9588,
+      "step": 40600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8728786423310924e-05,
+      "loss": 2.9534,
+      "step": 40700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8725584373999364e-05,
+      "loss": 2.9445,
+      "step": 40800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8722382324687804e-05,
+      "loss": 2.9338,
+      "step": 40900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8719180275376243e-05,
+      "loss": 2.98,
+      "step": 41000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.671525001525879,
+      "eval_runtime": 175.1529,
+      "eval_samples_per_second": 57.093,
+      "eval_steps_per_second": 3.568,
+      "step": 41000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.871597822606468e-05,
+      "loss": 2.952,
+      "step": 41100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.871277617675312e-05,
+      "loss": 2.9436,
+      "step": 41200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.870957412744156e-05,
+      "loss": 2.9371,
+      "step": 41300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.870637207813001e-05,
+      "loss": 2.9551,
+      "step": 41400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.870317002881845e-05,
+      "loss": 2.938,
+      "step": 41500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.869996797950689e-05,
+      "loss": 2.9563,
+      "step": 41600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.869676593019533e-05,
+      "loss": 2.9677,
+      "step": 41700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.869356388088377e-05,
+      "loss": 2.9425,
+      "step": 41800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.869036183157221e-05,
+      "loss": 2.9733,
+      "step": 41900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.868715978226065e-05,
+      "loss": 2.9384,
+      "step": 42000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6695194244384766,
+      "eval_runtime": 177.5003,
+      "eval_samples_per_second": 56.338,
+      "eval_steps_per_second": 3.521,
+      "step": 42000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.868395773294909e-05,
+      "loss": 2.9512,
+      "step": 42100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.868075568363753e-05,
+      "loss": 2.9533,
+      "step": 42200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8677553634325974e-05,
+      "loss": 2.9451,
+      "step": 42300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.867435158501441e-05,
+      "loss": 2.9512,
+      "step": 42400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.867114953570285e-05,
+      "loss": 2.953,
+      "step": 42500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.866794748639129e-05,
+      "loss": 2.9541,
+      "step": 42600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.866474543707973e-05,
+      "loss": 2.9516,
+      "step": 42700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.866154338776817e-05,
+      "loss": 2.9462,
+      "step": 42800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.865834133845661e-05,
+      "loss": 2.9565,
+      "step": 42900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.865513928914506e-05,
+      "loss": 2.9434,
+      "step": 43000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.663595676422119,
+      "eval_runtime": 177.606,
+      "eval_samples_per_second": 56.304,
+      "eval_steps_per_second": 3.519,
+      "step": 43000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.86519372398335e-05,
+      "loss": 2.9467,
+      "step": 43100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.864873519052194e-05,
+      "loss": 2.9429,
+      "step": 43200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.864553314121038e-05,
+      "loss": 2.9387,
+      "step": 43300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.864233109189882e-05,
+      "loss": 2.9609,
+      "step": 43400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.863912904258726e-05,
+      "loss": 2.9611,
+      "step": 43500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.86359269932757e-05,
+      "loss": 2.9362,
+      "step": 43600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8632724943964144e-05,
+      "loss": 2.9425,
+      "step": 43700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8629522894652576e-05,
+      "loss": 2.9494,
+      "step": 43800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.862632084534102e-05,
+      "loss": 2.9261,
+      "step": 43900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.862311879602946e-05,
+      "loss": 2.9369,
+      "step": 44000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6637465953826904,
+      "eval_runtime": 176.6283,
+      "eval_samples_per_second": 56.616,
+      "eval_steps_per_second": 3.539,
+      "step": 44000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.86199167467179e-05,
+      "loss": 2.9292,
+      "step": 44100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.861671469740634e-05,
+      "loss": 2.9375,
+      "step": 44200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.861351264809478e-05,
+      "loss": 2.9606,
+      "step": 44300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.861031059878323e-05,
+      "loss": 2.9337,
+      "step": 44400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.860710854947166e-05,
+      "loss": 2.9415,
+      "step": 44500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.860390650016011e-05,
+      "loss": 2.9612,
+      "step": 44600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.860070445084855e-05,
+      "loss": 2.9257,
+      "step": 44700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.859750240153699e-05,
+      "loss": 2.9384,
+      "step": 44800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.859430035222543e-05,
+      "loss": 2.9426,
+      "step": 44900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.859109830291387e-05,
+      "loss": 2.9349,
+      "step": 45000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6627140045166016,
+      "eval_runtime": 176.793,
+      "eval_samples_per_second": 56.563,
+      "eval_steps_per_second": 3.535,
+      "step": 45000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.858789625360231e-05,
+      "loss": 2.9307,
+      "step": 45100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8584694204290746e-05,
+      "loss": 2.9294,
+      "step": 45200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.858149215497919e-05,
+      "loss": 2.921,
+      "step": 45300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8578290105667626e-05,
+      "loss": 2.9285,
+      "step": 45400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.857508805635607e-05,
+      "loss": 2.9419,
+      "step": 45500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.857188600704451e-05,
+      "loss": 2.9245,
+      "step": 45600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.856868395773295e-05,
+      "loss": 2.9332,
+      "step": 45700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.856548190842139e-05,
+      "loss": 2.9201,
+      "step": 45800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.856227985910983e-05,
+      "loss": 2.9444,
+      "step": 45900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.855907780979828e-05,
+      "loss": 2.9321,
+      "step": 46000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6629557609558105,
+      "eval_runtime": 177.2252,
+      "eval_samples_per_second": 56.425,
+      "eval_steps_per_second": 3.527,
+      "step": 46000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.855587576048671e-05,
+      "loss": 2.9485,
+      "step": 46100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.855267371117516e-05,
+      "loss": 2.9094,
+      "step": 46200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.854947166186359e-05,
+      "loss": 2.9352,
+      "step": 46300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.854626961255204e-05,
+      "loss": 2.9278,
+      "step": 46400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8543067563240477e-05,
+      "loss": 2.9452,
+      "step": 46500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8539865513928916e-05,
+      "loss": 2.9558,
+      "step": 46600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.853666346461736e-05,
+      "loss": 2.9374,
+      "step": 46700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8533461415305796e-05,
+      "loss": 2.9263,
+      "step": 46800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.853025936599424e-05,
+      "loss": 2.9426,
+      "step": 46900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8527057316682675e-05,
+      "loss": 2.931,
+      "step": 47000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6591949462890625,
+      "eval_runtime": 175.6051,
+      "eval_samples_per_second": 56.946,
+      "eval_steps_per_second": 3.559,
+      "step": 47000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.852385526737112e-05,
+      "loss": 2.9286,
+      "step": 47100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.852065321805956e-05,
+      "loss": 2.9236,
+      "step": 47200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8517451168748e-05,
+      "loss": 2.9303,
+      "step": 47300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.851424911943644e-05,
+      "loss": 2.9203,
+      "step": 47400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.851104707012488e-05,
+      "loss": 2.9563,
+      "step": 47500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.850784502081333e-05,
+      "loss": 2.9176,
+      "step": 47600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.850464297150176e-05,
+      "loss": 2.9453,
+      "step": 47700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.850144092219021e-05,
+      "loss": 2.9437,
+      "step": 47800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.849823887287864e-05,
+      "loss": 2.9204,
+      "step": 47900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8495036823567086e-05,
+      "loss": 2.9205,
+      "step": 48000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6586101055145264,
+      "eval_runtime": 178.3108,
+      "eval_samples_per_second": 56.082,
+      "eval_steps_per_second": 3.505,
+      "step": 48000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8491834774255526e-05,
+      "loss": 2.9378,
+      "step": 48100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8488632724943966e-05,
+      "loss": 2.9229,
+      "step": 48200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.848543067563241e-05,
+      "loss": 2.9282,
+      "step": 48300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8482228626320845e-05,
+      "loss": 2.9312,
+      "step": 48400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.847902657700929e-05,
+      "loss": 2.9039,
+      "step": 48500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8475824527697725e-05,
+      "loss": 2.9361,
+      "step": 48600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.847262247838617e-05,
+      "loss": 2.9241,
+      "step": 48700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.846942042907461e-05,
+      "loss": 2.9296,
+      "step": 48800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.846621837976305e-05,
+      "loss": 2.9315,
+      "step": 48900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.84630163304515e-05,
+      "loss": 2.9147,
+      "step": 49000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.653660535812378,
+      "eval_runtime": 175.1839,
+      "eval_samples_per_second": 57.083,
+      "eval_steps_per_second": 3.568,
+      "step": 49000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.845981428113993e-05,
+      "loss": 2.9236,
+      "step": 49100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8456612231828377e-05,
+      "loss": 2.9235,
+      "step": 49200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.845341018251681e-05,
+      "loss": 2.926,
+      "step": 49300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8450208133205256e-05,
+      "loss": 2.8989,
+      "step": 49400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.844700608389369e-05,
+      "loss": 2.9362,
+      "step": 49500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8443804034582136e-05,
+      "loss": 2.9101,
+      "step": 49600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8440601985270575e-05,
+      "loss": 2.944,
+      "step": 49700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8437399935959015e-05,
+      "loss": 2.9158,
+      "step": 49800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.843419788664746e-05,
+      "loss": 2.9221,
+      "step": 49900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8430995837335894e-05,
+      "loss": 2.9005,
+      "step": 50000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6547129154205322,
+      "eval_runtime": 178.6492,
+      "eval_samples_per_second": 55.976,
+      "eval_steps_per_second": 3.498,
+      "step": 50000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.842779378802434e-05,
+      "loss": 2.9158,
+      "step": 50100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8424591738712774e-05,
+      "loss": 2.9161,
+      "step": 50200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.842138968940122e-05,
+      "loss": 2.9282,
+      "step": 50300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.841818764008966e-05,
+      "loss": 2.9058,
+      "step": 50400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.84149855907781e-05,
+      "loss": 2.9191,
+      "step": 50500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8411783541466546e-05,
+      "loss": 2.9172,
+      "step": 50600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.840858149215498e-05,
+      "loss": 2.8963,
+      "step": 50700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8405379442843426e-05,
+      "loss": 2.9345,
+      "step": 50800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.840217739353186e-05,
+      "loss": 2.9197,
+      "step": 50900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8398975344220305e-05,
+      "loss": 2.9018,
+      "step": 51000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.6506595611572266,
+      "eval_runtime": 178.1968,
+      "eval_samples_per_second": 56.118,
+      "eval_steps_per_second": 3.507,
+      "step": 51000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.839577329490874e-05,
+      "loss": 2.9245,
+      "step": 51100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8392571245597185e-05,
+      "loss": 2.92,
+      "step": 51200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8389369196285625e-05,
+      "loss": 2.9326,
+      "step": 51300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8386167146974064e-05,
+      "loss": 2.9369,
+      "step": 51400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.838296509766251e-05,
+      "loss": 2.8936,
+      "step": 51500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8379763048350944e-05,
+      "loss": 2.9116,
+      "step": 51600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.837656099903939e-05,
+      "loss": 2.9032,
+      "step": 51700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.837335894972782e-05,
+      "loss": 2.926,
+      "step": 51800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.837015690041627e-05,
+      "loss": 2.9204,
+      "step": 51900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.836695485110471e-05,
+      "loss": 2.9144,
+      "step": 52000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.64819073677063,
+      "eval_runtime": 175.6308,
+      "eval_samples_per_second": 56.938,
+      "eval_steps_per_second": 3.559,
+      "step": 52000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.836375280179315e-05,
+      "loss": 2.9234,
+      "step": 52100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8360550752481596e-05,
+      "loss": 2.9201,
+      "step": 52200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.835734870317003e-05,
+      "loss": 2.8795,
+      "step": 52300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8354146653858475e-05,
+      "loss": 2.9024,
+      "step": 52400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.835094460454691e-05,
+      "loss": 2.923,
+      "step": 52500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8347742555235355e-05,
+      "loss": 2.9094,
+      "step": 52600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.834454050592379e-05,
+      "loss": 2.9084,
+      "step": 52700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8341338456612234e-05,
+      "loss": 2.9046,
+      "step": 52800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8338136407300674e-05,
+      "loss": 2.897,
+      "step": 52900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8334934357989114e-05,
+      "loss": 2.8898,
+      "step": 53000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.645299196243286,
+      "eval_runtime": 177.085,
+      "eval_samples_per_second": 56.47,
+      "eval_steps_per_second": 3.529,
+      "step": 53000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.833173230867756e-05,
+      "loss": 2.9105,
+      "step": 53100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.832853025936599e-05,
+      "loss": 2.8876,
+      "step": 53200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.832532821005444e-05,
+      "loss": 2.917,
+      "step": 53300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.832212616074287e-05,
+      "loss": 2.8781,
+      "step": 53400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.831892411143132e-05,
+      "loss": 2.8958,
+      "step": 53500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.831572206211976e-05,
+      "loss": 2.9015,
+      "step": 53600
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.83125200128082e-05,
+      "loss": 2.8878,
+      "step": 53700
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8309317963496645e-05,
+      "loss": 2.9211,
+      "step": 53800
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.830611591418508e-05,
+      "loss": 2.9124,
+      "step": 53900
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8302913864873525e-05,
+      "loss": 2.9082,
+      "step": 54000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 2.64392352104187,
+      "eval_runtime": 177.3707,
+      "eval_samples_per_second": 56.379,
+      "eval_steps_per_second": 3.524,
+      "step": 54000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.829971181556196e-05,
+      "loss": 2.926,
+      "step": 54100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8296509766250404e-05,
+      "loss": 2.8951,
+      "step": 54200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8293307716938844e-05,
+      "loss": 2.901,
+      "step": 54300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.8290105667627284e-05,
+      "loss": 2.8796,
+      "step": 54400
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.828690361831572e-05,
+      "loss": 2.8966,
+      "step": 54500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.828370156900416e-05,
+      "loss": 2.9169,
+      "step": 54600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.828049951969261e-05,
+      "loss": 2.8987,
+      "step": 54700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.827729747038104e-05,
+      "loss": 2.8955,
+      "step": 54800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.827409542106949e-05,
+      "loss": 2.9023,
+      "step": 54900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.827089337175792e-05,
+      "loss": 2.9066,
+      "step": 55000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.642181634902954,
+      "eval_runtime": 176.3672,
+      "eval_samples_per_second": 56.7,
+      "eval_steps_per_second": 3.544,
+      "step": 55000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.826769132244637e-05,
+      "loss": 2.9017,
+      "step": 55100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.826448927313481e-05,
+      "loss": 2.9021,
+      "step": 55200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.826128722382325e-05,
+      "loss": 2.8813,
+      "step": 55300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8258085174511695e-05,
+      "loss": 2.9016,
+      "step": 55400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.825488312520013e-05,
+      "loss": 2.902,
+      "step": 55500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8251681075888574e-05,
+      "loss": 2.8913,
+      "step": 55600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.824847902657701e-05,
+      "loss": 2.8846,
+      "step": 55700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8245276977265454e-05,
+      "loss": 2.9166,
+      "step": 55800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.824207492795389e-05,
+      "loss": 2.897,
+      "step": 55900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.823887287864233e-05,
+      "loss": 2.8851,
+      "step": 56000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6408231258392334,
+      "eval_runtime": 178.371,
+      "eval_samples_per_second": 56.063,
+      "eval_steps_per_second": 3.504,
+      "step": 56000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.823567082933077e-05,
+      "loss": 2.8904,
+      "step": 56100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.823246878001921e-05,
+      "loss": 2.9088,
+      "step": 56200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.822926673070766e-05,
+      "loss": 2.9073,
+      "step": 56300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.822606468139609e-05,
+      "loss": 2.8893,
+      "step": 56400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.822286263208454e-05,
+      "loss": 2.8898,
+      "step": 56500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.821966058277298e-05,
+      "loss": 2.9206,
+      "step": 56600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.821645853346142e-05,
+      "loss": 2.9178,
+      "step": 56700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.821325648414986e-05,
+      "loss": 2.8975,
+      "step": 56800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.82100544348383e-05,
+      "loss": 2.9013,
+      "step": 56900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8206852385526744e-05,
+      "loss": 2.8921,
+      "step": 57000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6402812004089355,
+      "eval_runtime": 176.0762,
+      "eval_samples_per_second": 56.794,
+      "eval_steps_per_second": 3.55,
+      "step": 57000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.820365033621518e-05,
+      "loss": 2.8888,
+      "step": 57100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8200448286903623e-05,
+      "loss": 2.8943,
+      "step": 57200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8197246237592056e-05,
+      "loss": 2.9095,
+      "step": 57300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.81940441882805e-05,
+      "loss": 2.8993,
+      "step": 57400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.819084213896894e-05,
+      "loss": 2.886,
+      "step": 57500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.818764008965738e-05,
+      "loss": 2.8974,
+      "step": 57600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.818443804034582e-05,
+      "loss": 2.8738,
+      "step": 57700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.818123599103426e-05,
+      "loss": 2.9148,
+      "step": 57800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.817803394172271e-05,
+      "loss": 2.8871,
+      "step": 57900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.817483189241114e-05,
+      "loss": 2.8949,
+      "step": 58000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6379764080047607,
+      "eval_runtime": 178.1605,
+      "eval_samples_per_second": 56.129,
+      "eval_steps_per_second": 3.508,
+      "step": 58000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.817162984309959e-05,
+      "loss": 2.8779,
+      "step": 58100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.816842779378803e-05,
+      "loss": 2.877,
+      "step": 58200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.816522574447647e-05,
+      "loss": 2.8852,
+      "step": 58300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.816202369516491e-05,
+      "loss": 2.8913,
+      "step": 58400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.815882164585335e-05,
+      "loss": 2.8927,
+      "step": 58500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.815561959654179e-05,
+      "loss": 2.9023,
+      "step": 58600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8152417547230226e-05,
+      "loss": 2.8747,
+      "step": 58700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.814921549791867e-05,
+      "loss": 2.925,
+      "step": 58800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.814601344860711e-05,
+      "loss": 2.8965,
+      "step": 58900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.814281139929555e-05,
+      "loss": 2.8893,
+      "step": 59000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6361653804779053,
+      "eval_runtime": 176.4963,
+      "eval_samples_per_second": 56.658,
+      "eval_steps_per_second": 3.541,
+      "step": 59000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.813960934998399e-05,
+      "loss": 2.8837,
+      "step": 59100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.813640730067243e-05,
+      "loss": 2.8872,
+      "step": 59200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.813320525136087e-05,
+      "loss": 2.8987,
+      "step": 59300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.813000320204931e-05,
+      "loss": 2.8938,
+      "step": 59400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.812680115273776e-05,
+      "loss": 2.887,
+      "step": 59500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.812359910342619e-05,
+      "loss": 2.8972,
+      "step": 59600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.812039705411464e-05,
+      "loss": 2.8788,
+      "step": 59700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.811719500480308e-05,
+      "loss": 2.8931,
+      "step": 59800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.811399295549152e-05,
+      "loss": 2.8809,
+      "step": 59900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8110790906179956e-05,
+      "loss": 2.895,
+      "step": 60000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.636247158050537,
+      "eval_runtime": 176.1455,
+      "eval_samples_per_second": 56.771,
+      "eval_steps_per_second": 3.548,
+      "step": 60000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8107588856868396e-05,
+      "loss": 2.8971,
+      "step": 60100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.810438680755684e-05,
+      "loss": 2.8754,
+      "step": 60200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8101184758245276e-05,
+      "loss": 2.9116,
+      "step": 60300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.809798270893372e-05,
+      "loss": 2.8929,
+      "step": 60400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.809478065962216e-05,
+      "loss": 2.8687,
+      "step": 60500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.80915786103106e-05,
+      "loss": 2.8883,
+      "step": 60600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.808837656099904e-05,
+      "loss": 2.8928,
+      "step": 60700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.808517451168748e-05,
+      "loss": 2.8977,
+      "step": 60800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.808197246237592e-05,
+      "loss": 2.8938,
+      "step": 60900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.807877041306436e-05,
+      "loss": 2.8765,
+      "step": 61000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6320600509643555,
+      "eval_runtime": 177.2981,
+      "eval_samples_per_second": 56.402,
+      "eval_steps_per_second": 3.525,
+      "step": 61000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.807556836375281e-05,
+      "loss": 2.8716,
+      "step": 61100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.807236631444125e-05,
+      "loss": 2.8953,
+      "step": 61200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8069164265129687e-05,
+      "loss": 2.8848,
+      "step": 61300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8065962215818126e-05,
+      "loss": 2.892,
+      "step": 61400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8062760166506566e-05,
+      "loss": 2.8878,
+      "step": 61500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8059558117195006e-05,
+      "loss": 2.8692,
+      "step": 61600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8056356067883446e-05,
+      "loss": 2.8683,
+      "step": 61700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.805315401857189e-05,
+      "loss": 2.8776,
+      "step": 61800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8049951969260325e-05,
+      "loss": 2.8778,
+      "step": 61900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.804674991994877e-05,
+      "loss": 2.8915,
+      "step": 62000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.630981206893921,
+      "eval_runtime": 176.4905,
+      "eval_samples_per_second": 56.66,
+      "eval_steps_per_second": 3.541,
+      "step": 62000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.804354787063721e-05,
+      "loss": 2.8944,
+      "step": 62100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.804034582132565e-05,
+      "loss": 2.8721,
+      "step": 62200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.803714377201409e-05,
+      "loss": 2.8661,
+      "step": 62300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.803394172270253e-05,
+      "loss": 2.8959,
+      "step": 62400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.803073967339097e-05,
+      "loss": 2.8752,
+      "step": 62500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.802753762407941e-05,
+      "loss": 2.8858,
+      "step": 62600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8024335574767856e-05,
+      "loss": 2.8771,
+      "step": 62700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8021133525456296e-05,
+      "loss": 2.8746,
+      "step": 62800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8017931476144736e-05,
+      "loss": 2.8726,
+      "step": 62900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8014729426833176e-05,
+      "loss": 2.8841,
+      "step": 63000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.631925106048584,
+      "eval_runtime": 178.91,
+      "eval_samples_per_second": 55.894,
+      "eval_steps_per_second": 3.493,
+      "step": 63000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8011527377521615e-05,
+      "loss": 2.8828,
+      "step": 63100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8008325328210055e-05,
+      "loss": 2.8462,
+      "step": 63200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8005123278898495e-05,
+      "loss": 2.8835,
+      "step": 63300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8001921229586935e-05,
+      "loss": 2.8805,
+      "step": 63400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.799871918027538e-05,
+      "loss": 2.8842,
+      "step": 63500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.799551713096382e-05,
+      "loss": 2.8914,
+      "step": 63600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.799231508165226e-05,
+      "loss": 2.8578,
+      "step": 63700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.79891130323407e-05,
+      "loss": 2.8729,
+      "step": 63800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.798591098302914e-05,
+      "loss": 2.8858,
+      "step": 63900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.798270893371758e-05,
+      "loss": 2.8701,
+      "step": 64000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6272077560424805,
+      "eval_runtime": 176.6554,
+      "eval_samples_per_second": 56.607,
+      "eval_steps_per_second": 3.538,
+      "step": 64000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.797950688440602e-05,
+      "loss": 2.8801,
+      "step": 64100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7976304835094466e-05,
+      "loss": 2.9025,
+      "step": 64200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7973102785782906e-05,
+      "loss": 2.8743,
+      "step": 64300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7969900736471346e-05,
+      "loss": 2.8923,
+      "step": 64400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7966698687159785e-05,
+      "loss": 2.8944,
+      "step": 64500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7963496637848225e-05,
+      "loss": 2.887,
+      "step": 64600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7960294588536665e-05,
+      "loss": 2.8828,
+      "step": 64700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7957092539225105e-05,
+      "loss": 2.8754,
+      "step": 64800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7953890489913544e-05,
+      "loss": 2.8791,
+      "step": 64900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7950688440601984e-05,
+      "loss": 2.8407,
+      "step": 65000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6278929710388184,
+      "eval_runtime": 177.1352,
+      "eval_samples_per_second": 56.454,
+      "eval_steps_per_second": 3.528,
+      "step": 65000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.794748639129043e-05,
+      "loss": 2.8656,
+      "step": 65100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.794428434197887e-05,
+      "loss": 2.8675,
+      "step": 65200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.794108229266731e-05,
+      "loss": 2.8924,
+      "step": 65300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.793788024335575e-05,
+      "loss": 2.8682,
+      "step": 65400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.793467819404419e-05,
+      "loss": 2.883,
+      "step": 65500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.793147614473263e-05,
+      "loss": 2.8548,
+      "step": 65600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.792827409542107e-05,
+      "loss": 2.8845,
+      "step": 65700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7925072046109515e-05,
+      "loss": 2.8869,
+      "step": 65800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7921869996797955e-05,
+      "loss": 2.857,
+      "step": 65900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7918667947486395e-05,
+      "loss": 2.8659,
+      "step": 66000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.6252381801605225,
+      "eval_runtime": 178.5365,
+      "eval_samples_per_second": 56.011,
+      "eval_steps_per_second": 3.501,
+      "step": 66000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7915465898174835e-05,
+      "loss": 2.8654,
+      "step": 66100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7912263848863274e-05,
+      "loss": 2.87,
+      "step": 66200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7909061799551714e-05,
+      "loss": 2.8696,
+      "step": 66300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7905859750240154e-05,
+      "loss": 2.87,
+      "step": 66400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.79026577009286e-05,
+      "loss": 2.8805,
+      "step": 66500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7899455651617033e-05,
+      "loss": 2.8696,
+      "step": 66600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.789625360230548e-05,
+      "loss": 2.8675,
+      "step": 66700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.789305155299392e-05,
+      "loss": 2.8673,
+      "step": 66800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.788984950368236e-05,
+      "loss": 2.8822,
+      "step": 66900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.78866474543708e-05,
+      "loss": 2.8722,
+      "step": 67000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.626574993133545,
+      "eval_runtime": 175.2826,
+      "eval_samples_per_second": 57.051,
+      "eval_steps_per_second": 3.566,
+      "step": 67000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.788344540505924e-05,
+      "loss": 2.8481,
+      "step": 67100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.788024335574768e-05,
+      "loss": 2.8707,
+      "step": 67200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.787704130643612e-05,
+      "loss": 2.8761,
+      "step": 67300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7873839257124565e-05,
+      "loss": 2.8592,
+      "step": 67400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7870637207813005e-05,
+      "loss": 2.8697,
+      "step": 67500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7867435158501444e-05,
+      "loss": 2.8961,
+      "step": 67600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7864233109189884e-05,
+      "loss": 2.8615,
+      "step": 67700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7861031059878324e-05,
+      "loss": 2.8696,
+      "step": 67800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7857829010566764e-05,
+      "loss": 2.8787,
+      "step": 67900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.78546269612552e-05,
+      "loss": 2.8539,
+      "step": 68000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.624630928039551,
+      "eval_runtime": 179.2846,
+      "eval_samples_per_second": 55.777,
+      "eval_steps_per_second": 3.486,
+      "step": 68000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.785142491194365e-05,
+      "loss": 2.8506,
+      "step": 68100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.784822286263208e-05,
+      "loss": 2.8606,
+      "step": 68200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.784502081332053e-05,
+      "loss": 2.8456,
+      "step": 68300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.784181876400897e-05,
+      "loss": 2.8524,
+      "step": 68400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.783861671469741e-05,
+      "loss": 2.8596,
+      "step": 68500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.783541466538585e-05,
+      "loss": 2.8607,
+      "step": 68600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.783221261607429e-05,
+      "loss": 2.8647,
+      "step": 68700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7829010566762735e-05,
+      "loss": 2.8447,
+      "step": 68800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.782580851745117e-05,
+      "loss": 2.8862,
+      "step": 68900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7822606468139614e-05,
+      "loss": 2.8641,
+      "step": 69000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.623765468597412,
+      "eval_runtime": 177.1886,
+      "eval_samples_per_second": 56.437,
+      "eval_steps_per_second": 3.527,
+      "step": 69000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7819404418828054e-05,
+      "loss": 2.8487,
+      "step": 69100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7816202369516494e-05,
+      "loss": 2.8696,
+      "step": 69200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7813000320204933e-05,
+      "loss": 2.8499,
+      "step": 69300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.780979827089337e-05,
+      "loss": 2.8757,
+      "step": 69400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.780659622158181e-05,
+      "loss": 2.8534,
+      "step": 69500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.780339417227025e-05,
+      "loss": 2.8931,
+      "step": 69600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.78001921229587e-05,
+      "loss": 2.8608,
+      "step": 69700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.779699007364713e-05,
+      "loss": 2.8753,
+      "step": 69800
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.779378802433558e-05,
+      "loss": 2.8602,
+      "step": 69900
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.779058597502402e-05,
+      "loss": 2.8386,
+      "step": 70000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 2.623189687728882,
+      "eval_runtime": 178.3982,
+      "eval_samples_per_second": 56.054,
+      "eval_steps_per_second": 3.503,
+      "step": 70000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.778738392571246e-05,
+      "loss": 2.8434,
+      "step": 70100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.77841818764009e-05,
+      "loss": 2.8547,
+      "step": 70200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.778097982708934e-05,
+      "loss": 2.8507,
+      "step": 70300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 2.83,
+      "step": 70400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.777457572846622e-05,
+      "loss": 2.8491,
+      "step": 70500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7771373679154664e-05,
+      "loss": 2.8496,
+      "step": 70600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.77681716298431e-05,
+      "loss": 2.8441,
+      "step": 70700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.776496958053154e-05,
+      "loss": 2.8496,
+      "step": 70800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.776176753121998e-05,
+      "loss": 2.8633,
+      "step": 70900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.775856548190842e-05,
+      "loss": 2.8588,
+      "step": 71000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.62199068069458,
+      "eval_runtime": 177.949,
+      "eval_samples_per_second": 56.196,
+      "eval_steps_per_second": 3.512,
+      "step": 71000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.775536343259687e-05,
+      "loss": 2.8538,
+      "step": 71100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.77521613832853e-05,
+      "loss": 2.8511,
+      "step": 71200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.774895933397375e-05,
+      "loss": 2.8625,
+      "step": 71300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.774575728466218e-05,
+      "loss": 2.861,
+      "step": 71400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.774255523535063e-05,
+      "loss": 2.8455,
+      "step": 71500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.773935318603907e-05,
+      "loss": 2.8341,
+      "step": 71600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.773615113672751e-05,
+      "loss": 2.8646,
+      "step": 71700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.773294908741595e-05,
+      "loss": 2.8435,
+      "step": 71800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.772974703810439e-05,
+      "loss": 2.832,
+      "step": 71900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7726544988792834e-05,
+      "loss": 2.8626,
+      "step": 72000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6206774711608887,
+      "eval_runtime": 179.2437,
+      "eval_samples_per_second": 55.79,
+      "eval_steps_per_second": 3.487,
+      "step": 72000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7723342939481266e-05,
+      "loss": 2.8369,
+      "step": 72100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.772014089016971e-05,
+      "loss": 2.8413,
+      "step": 72200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.771693884085815e-05,
+      "loss": 2.8481,
+      "step": 72300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.771373679154659e-05,
+      "loss": 2.8515,
+      "step": 72400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.771053474223503e-05,
+      "loss": 2.836,
+      "step": 72500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.770733269292347e-05,
+      "loss": 2.8408,
+      "step": 72600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.770413064361192e-05,
+      "loss": 2.8476,
+      "step": 72700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.770092859430035e-05,
+      "loss": 2.8521,
+      "step": 72800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.76977265449888e-05,
+      "loss": 2.8486,
+      "step": 72900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.769452449567723e-05,
+      "loss": 2.8588,
+      "step": 73000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6182003021240234,
+      "eval_runtime": 179.3657,
+      "eval_samples_per_second": 55.752,
+      "eval_steps_per_second": 3.485,
+      "step": 73000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.769132244636568e-05,
+      "loss": 2.8372,
+      "step": 73100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.768812039705412e-05,
+      "loss": 2.8277,
+      "step": 73200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.768491834774256e-05,
+      "loss": 2.8451,
+      "step": 73300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7681716298431003e-05,
+      "loss": 2.8498,
+      "step": 73400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7678514249119436e-05,
+      "loss": 2.8222,
+      "step": 73500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.767531219980788e-05,
+      "loss": 2.8307,
+      "step": 73600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7672110150496316e-05,
+      "loss": 2.8404,
+      "step": 73700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.766890810118476e-05,
+      "loss": 2.8521,
+      "step": 73800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.76657060518732e-05,
+      "loss": 2.8389,
+      "step": 73900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.766250400256164e-05,
+      "loss": 2.8513,
+      "step": 74000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6168673038482666,
+      "eval_runtime": 179.3293,
+      "eval_samples_per_second": 55.763,
+      "eval_steps_per_second": 3.485,
+      "step": 74000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.765930195325009e-05,
+      "loss": 2.8545,
+      "step": 74100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.765609990393852e-05,
+      "loss": 2.8304,
+      "step": 74200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.765289785462697e-05,
+      "loss": 2.8522,
+      "step": 74300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.76496958053154e-05,
+      "loss": 2.8184,
+      "step": 74400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.764649375600385e-05,
+      "loss": 2.8588,
+      "step": 74500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.764329170669228e-05,
+      "loss": 2.8538,
+      "step": 74600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.764008965738073e-05,
+      "loss": 2.8357,
+      "step": 74700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7636887608069167e-05,
+      "loss": 2.824,
+      "step": 74800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7633685558757606e-05,
+      "loss": 2.8445,
+      "step": 74900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.763048350944605e-05,
+      "loss": 2.8508,
+      "step": 75000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6163156032562256,
+      "eval_runtime": 174.9884,
+      "eval_samples_per_second": 57.147,
+      "eval_steps_per_second": 3.572,
+      "step": 75000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7627281460134486e-05,
+      "loss": 2.828,
+      "step": 75100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.762407941082293e-05,
+      "loss": 2.8371,
+      "step": 75200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7620877361511365e-05,
+      "loss": 2.8336,
+      "step": 75300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.761767531219981e-05,
+      "loss": 2.8408,
+      "step": 75400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.761447326288825e-05,
+      "loss": 2.8385,
+      "step": 75500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.761127121357669e-05,
+      "loss": 2.8697,
+      "step": 75600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.760806916426514e-05,
+      "loss": 2.8561,
+      "step": 75700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.760486711495357e-05,
+      "loss": 2.8452,
+      "step": 75800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.760166506564202e-05,
+      "loss": 2.8632,
+      "step": 75900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.759846301633045e-05,
+      "loss": 2.8484,
+      "step": 76000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.615889072418213,
+      "eval_runtime": 175.9824,
+      "eval_samples_per_second": 56.824,
+      "eval_steps_per_second": 3.551,
+      "step": 76000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.75952609670189e-05,
+      "loss": 2.847,
+      "step": 76100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.759205891770733e-05,
+      "loss": 2.8438,
+      "step": 76200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7588856868395776e-05,
+      "loss": 2.8175,
+      "step": 76300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7585654819084216e-05,
+      "loss": 2.829,
+      "step": 76400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7582452769772656e-05,
+      "loss": 2.8611,
+      "step": 76500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.75792507204611e-05,
+      "loss": 2.851,
+      "step": 76600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7576048671149535e-05,
+      "loss": 2.8397,
+      "step": 76700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.757284662183798e-05,
+      "loss": 2.8409,
+      "step": 76800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7569644572526415e-05,
+      "loss": 2.833,
+      "step": 76900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.756644252321486e-05,
+      "loss": 2.8406,
+      "step": 77000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.613297700881958,
+      "eval_runtime": 177.3489,
+      "eval_samples_per_second": 56.386,
+      "eval_steps_per_second": 3.524,
+      "step": 77000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.75632404739033e-05,
+      "loss": 2.836,
+      "step": 77100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.756003842459174e-05,
+      "loss": 2.8227,
+      "step": 77200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.755683637528019e-05,
+      "loss": 2.8241,
+      "step": 77300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.755363432596862e-05,
+      "loss": 2.8277,
+      "step": 77400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7550432276657067e-05,
+      "loss": 2.8311,
+      "step": 77500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.75472302273455e-05,
+      "loss": 2.8333,
+      "step": 77600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7544028178033946e-05,
+      "loss": 2.8526,
+      "step": 77700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.754082612872238e-05,
+      "loss": 2.8425,
+      "step": 77800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7537624079410826e-05,
+      "loss": 2.8275,
+      "step": 77900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7534422030099265e-05,
+      "loss": 2.8275,
+      "step": 78000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.611859083175659,
+      "eval_runtime": 176.8529,
+      "eval_samples_per_second": 56.544,
+      "eval_steps_per_second": 3.534,
+      "step": 78000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7531219980787705e-05,
+      "loss": 2.8123,
+      "step": 78100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.752801793147615e-05,
+      "loss": 2.842,
+      "step": 78200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7524815882164584e-05,
+      "loss": 2.8515,
+      "step": 78300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.752161383285303e-05,
+      "loss": 2.8293,
+      "step": 78400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7518411783541464e-05,
+      "loss": 2.8288,
+      "step": 78500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.751520973422991e-05,
+      "loss": 2.8517,
+      "step": 78600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.751200768491835e-05,
+      "loss": 2.8363,
+      "step": 78700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.750880563560679e-05,
+      "loss": 2.8586,
+      "step": 78800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7505603586295236e-05,
+      "loss": 2.8319,
+      "step": 78900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.750240153698367e-05,
+      "loss": 2.8237,
+      "step": 79000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.610048532485962,
+      "eval_runtime": 177.9401,
+      "eval_samples_per_second": 56.199,
+      "eval_steps_per_second": 3.512,
+      "step": 79000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7499199487672116e-05,
+      "loss": 2.8294,
+      "step": 79100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.749599743836055e-05,
+      "loss": 2.8175,
+      "step": 79200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7492795389048995e-05,
+      "loss": 2.8375,
+      "step": 79300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.748959333973743e-05,
+      "loss": 2.8302,
+      "step": 79400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7486391290425875e-05,
+      "loss": 2.8193,
+      "step": 79500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7483189241114315e-05,
+      "loss": 2.8389,
+      "step": 79600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7479987191802754e-05,
+      "loss": 2.8373,
+      "step": 79700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.74767851424912e-05,
+      "loss": 2.8473,
+      "step": 79800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7473583093179634e-05,
+      "loss": 2.8022,
+      "step": 79900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.747038104386808e-05,
+      "loss": 2.8151,
+      "step": 80000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6109461784362793,
+      "eval_runtime": 179.9943,
+      "eval_samples_per_second": 55.557,
+      "eval_steps_per_second": 3.472,
+      "step": 80000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.746717899455651e-05,
+      "loss": 2.8428,
+      "step": 80100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.746397694524496e-05,
+      "loss": 2.8367,
+      "step": 80200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.74607748959334e-05,
+      "loss": 2.8256,
+      "step": 80300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.745757284662184e-05,
+      "loss": 2.8322,
+      "step": 80400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7454370797310286e-05,
+      "loss": 2.8296,
+      "step": 80500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.745116874799872e-05,
+      "loss": 2.8231,
+      "step": 80600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7447966698687165e-05,
+      "loss": 2.8306,
+      "step": 80700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.74447646493756e-05,
+      "loss": 2.8411,
+      "step": 80800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7441562600064045e-05,
+      "loss": 2.8214,
+      "step": 80900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7438360550752485e-05,
+      "loss": 2.8316,
+      "step": 81000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6093032360076904,
+      "eval_runtime": 180.9743,
+      "eval_samples_per_second": 55.256,
+      "eval_steps_per_second": 3.454,
+      "step": 81000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7435158501440924e-05,
+      "loss": 2.8423,
+      "step": 81100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7431956452129364e-05,
+      "loss": 2.8475,
+      "step": 81200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7428754402817804e-05,
+      "loss": 2.8288,
+      "step": 81300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.742555235350625e-05,
+      "loss": 2.8268,
+      "step": 81400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.742235030419468e-05,
+      "loss": 2.8296,
+      "step": 81500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.741914825488313e-05,
+      "loss": 2.8115,
+      "step": 81600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.741594620557156e-05,
+      "loss": 2.8189,
+      "step": 81700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.741274415626001e-05,
+      "loss": 2.826,
+      "step": 81800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.740954210694845e-05,
+      "loss": 2.8234,
+      "step": 81900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.740634005763689e-05,
+      "loss": 2.8323,
+      "step": 82000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6061840057373047,
+      "eval_runtime": 176.2652,
+      "eval_samples_per_second": 56.733,
+      "eval_steps_per_second": 3.546,
+      "step": 82000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.740313800832533e-05,
+      "loss": 2.8221,
+      "step": 82100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.739993595901377e-05,
+      "loss": 2.8164,
+      "step": 82200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7396733909702215e-05,
+      "loss": 2.8348,
+      "step": 82300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.739353186039065e-05,
+      "loss": 2.8126,
+      "step": 82400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7390329811079094e-05,
+      "loss": 2.8422,
+      "step": 82500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7387127761767534e-05,
+      "loss": 2.8054,
+      "step": 82600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7383925712455974e-05,
+      "loss": 2.8152,
+      "step": 82700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.738072366314441e-05,
+      "loss": 2.8113,
+      "step": 82800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.737752161383285e-05,
+      "loss": 2.8288,
+      "step": 82900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.73743195645213e-05,
+      "loss": 2.8339,
+      "step": 83000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6048617362976074,
+      "eval_runtime": 180.0118,
+      "eval_samples_per_second": 55.552,
+      "eval_steps_per_second": 3.472,
+      "step": 83000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.737111751520973e-05,
+      "loss": 2.8217,
+      "step": 83100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.736791546589818e-05,
+      "loss": 2.805,
+      "step": 83200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.736471341658662e-05,
+      "loss": 2.8304,
+      "step": 83300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.736151136727506e-05,
+      "loss": 2.8284,
+      "step": 83400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.73583093179635e-05,
+      "loss": 2.8139,
+      "step": 83500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.735510726865194e-05,
+      "loss": 2.8407,
+      "step": 83600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.735190521934038e-05,
+      "loss": 2.8052,
+      "step": 83700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.734870317002882e-05,
+      "loss": 2.8452,
+      "step": 83800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7345501120717264e-05,
+      "loss": 2.8436,
+      "step": 83900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7342299071405704e-05,
+      "loss": 2.8197,
+      "step": 84000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.6033012866973877,
+      "eval_runtime": 177.5805,
+      "eval_samples_per_second": 56.312,
+      "eval_steps_per_second": 3.52,
+      "step": 84000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7339097022094144e-05,
+      "loss": 2.8055,
+      "step": 84100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.733589497278258e-05,
+      "loss": 2.8088,
+      "step": 84200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.733269292347102e-05,
+      "loss": 2.8306,
+      "step": 84300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.732949087415946e-05,
+      "loss": 2.8089,
+      "step": 84400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.73262888248479e-05,
+      "loss": 2.7994,
+      "step": 84500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.732308677553635e-05,
+      "loss": 2.8545,
+      "step": 84600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.731988472622478e-05,
+      "loss": 2.8154,
+      "step": 84700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.731668267691323e-05,
+      "loss": 2.8359,
+      "step": 84800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.731348062760167e-05,
+      "loss": 2.8292,
+      "step": 84900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.731027857829011e-05,
+      "loss": 2.7941,
+      "step": 85000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 2.604931116104126,
+      "eval_runtime": 180.7023,
+      "eval_samples_per_second": 55.34,
+      "eval_steps_per_second": 3.459,
+      "step": 85000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.730707652897855e-05,
+      "loss": 2.8105,
+      "step": 85100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.730387447966699e-05,
+      "loss": 2.8295,
+      "step": 85200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.730067243035543e-05,
+      "loss": 2.7998,
+      "step": 85300
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.729747038104387e-05,
+      "loss": 2.821,
+      "step": 85400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.7294268331732313e-05,
+      "loss": 2.8151,
+      "step": 85500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.729106628242075e-05,
+      "loss": 2.8,
+      "step": 85600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.728786423310919e-05,
+      "loss": 2.8241,
+      "step": 85700
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.728466218379763e-05,
+      "loss": 2.813,
+      "step": 85800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.728146013448607e-05,
+      "loss": 2.8031,
+      "step": 85900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.727825808517451e-05,
+      "loss": 2.791,
+      "step": 86000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.6037590503692627,
+      "eval_runtime": 176.2434,
+      "eval_samples_per_second": 56.74,
+      "eval_steps_per_second": 3.546,
+      "step": 86000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.727505603586295e-05,
+      "loss": 2.7966,
+      "step": 86100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.72718539865514e-05,
+      "loss": 2.8253,
+      "step": 86200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.726865193723984e-05,
+      "loss": 2.7886,
+      "step": 86300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.726544988792828e-05,
+      "loss": 2.7999,
+      "step": 86400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.726224783861672e-05,
+      "loss": 2.8172,
+      "step": 86500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.725904578930516e-05,
+      "loss": 2.7939,
+      "step": 86600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.72558437399936e-05,
+      "loss": 2.8033,
+      "step": 86700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.725264169068204e-05,
+      "loss": 2.8201,
+      "step": 86800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7249439641370477e-05,
+      "loss": 2.8243,
+      "step": 86900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7246237592058916e-05,
+      "loss": 2.8016,
+      "step": 87000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.6006062030792236,
+      "eval_runtime": 179.333,
+      "eval_samples_per_second": 55.762,
+      "eval_steps_per_second": 3.485,
+      "step": 87000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.724303554274736e-05,
+      "loss": 2.8173,
+      "step": 87100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.72398334934358e-05,
+      "loss": 2.817,
+      "step": 87200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.723663144412424e-05,
+      "loss": 2.8145,
+      "step": 87300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.723342939481268e-05,
+      "loss": 2.7841,
+      "step": 87400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.723022734550112e-05,
+      "loss": 2.8264,
+      "step": 87500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.722702529618956e-05,
+      "loss": 2.8281,
+      "step": 87600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7223823246878e-05,
+      "loss": 2.8169,
+      "step": 87700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.722062119756645e-05,
+      "loss": 2.8085,
+      "step": 87800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.721741914825489e-05,
+      "loss": 2.8151,
+      "step": 87900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.721421709894333e-05,
+      "loss": 2.7913,
+      "step": 88000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5992484092712402,
+      "eval_runtime": 178.437,
+      "eval_samples_per_second": 56.042,
+      "eval_steps_per_second": 3.503,
+      "step": 88000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.721101504963177e-05,
+      "loss": 2.8022,
+      "step": 88100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.720781300032021e-05,
+      "loss": 2.7993,
+      "step": 88200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7204610951008646e-05,
+      "loss": 2.8054,
+      "step": 88300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7201408901697086e-05,
+      "loss": 2.8266,
+      "step": 88400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7198206852385526e-05,
+      "loss": 2.7973,
+      "step": 88500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.719500480307397e-05,
+      "loss": 2.7932,
+      "step": 88600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.719180275376241e-05,
+      "loss": 2.8047,
+      "step": 88700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.718860070445085e-05,
+      "loss": 2.8009,
+      "step": 88800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.718539865513929e-05,
+      "loss": 2.794,
+      "step": 88900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.718219660582773e-05,
+      "loss": 2.8174,
+      "step": 89000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5968127250671387,
+      "eval_runtime": 177.4518,
+      "eval_samples_per_second": 56.353,
+      "eval_steps_per_second": 3.522,
+      "step": 89000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.717899455651617e-05,
+      "loss": 2.8056,
+      "step": 89100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.717579250720461e-05,
+      "loss": 2.792,
+      "step": 89200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.717259045789305e-05,
+      "loss": 2.7907,
+      "step": 89300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.71693884085815e-05,
+      "loss": 2.8287,
+      "step": 89400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.716618635926994e-05,
+      "loss": 2.803,
+      "step": 89500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7162984309958377e-05,
+      "loss": 2.8202,
+      "step": 89600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7159782260646816e-05,
+      "loss": 2.8129,
+      "step": 89700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7156580211335256e-05,
+      "loss": 2.8159,
+      "step": 89800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7153378162023696e-05,
+      "loss": 2.8032,
+      "step": 89900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7150176112712136e-05,
+      "loss": 2.8074,
+      "step": 90000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.595768451690674,
+      "eval_runtime": 180.4275,
+      "eval_samples_per_second": 55.424,
+      "eval_steps_per_second": 3.464,
+      "step": 90000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7146974063400575e-05,
+      "loss": 2.8138,
+      "step": 90100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.714377201408902e-05,
+      "loss": 2.808,
+      "step": 90200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.714056996477746e-05,
+      "loss": 2.7893,
+      "step": 90300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.71373679154659e-05,
+      "loss": 2.7843,
+      "step": 90400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.713416586615434e-05,
+      "loss": 2.8119,
+      "step": 90500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.713096381684278e-05,
+      "loss": 2.794,
+      "step": 90600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.712776176753122e-05,
+      "loss": 2.8023,
+      "step": 90700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.712455971821966e-05,
+      "loss": 2.7732,
+      "step": 90800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.712135766890811e-05,
+      "loss": 2.7963,
+      "step": 90900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7118155619596546e-05,
+      "loss": 2.8083,
+      "step": 91000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5970373153686523,
+      "eval_runtime": 178.3132,
+      "eval_samples_per_second": 56.081,
+      "eval_steps_per_second": 3.505,
+      "step": 91000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7114953570284986e-05,
+      "loss": 2.7963,
+      "step": 91100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7111751520973426e-05,
+      "loss": 2.8041,
+      "step": 91200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7108549471661866e-05,
+      "loss": 2.7839,
+      "step": 91300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7105347422350305e-05,
+      "loss": 2.8125,
+      "step": 91400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7102145373038745e-05,
+      "loss": 2.8114,
+      "step": 91500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.709894332372719e-05,
+      "loss": 2.8269,
+      "step": 91600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7095741274415625e-05,
+      "loss": 2.8165,
+      "step": 91700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.709253922510407e-05,
+      "loss": 2.804,
+      "step": 91800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.708933717579251e-05,
+      "loss": 2.777,
+      "step": 91900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.708613512648095e-05,
+      "loss": 2.8046,
+      "step": 92000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5963518619537354,
+      "eval_runtime": 176.577,
+      "eval_samples_per_second": 56.633,
+      "eval_steps_per_second": 3.54,
+      "step": 92000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.708293307716939e-05,
+      "loss": 2.8146,
+      "step": 92100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.707973102785783e-05,
+      "loss": 2.8026,
+      "step": 92200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.707652897854627e-05,
+      "loss": 2.8044,
+      "step": 92300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.707332692923471e-05,
+      "loss": 2.8016,
+      "step": 92400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7070124879923156e-05,
+      "loss": 2.7925,
+      "step": 92500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7066922830611596e-05,
+      "loss": 2.8174,
+      "step": 92600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7063720781300036e-05,
+      "loss": 2.7998,
+      "step": 92700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7060518731988475e-05,
+      "loss": 2.8043,
+      "step": 92800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7057316682676915e-05,
+      "loss": 2.8038,
+      "step": 92900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7054114633365355e-05,
+      "loss": 2.7837,
+      "step": 93000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5965936183929443,
+      "eval_runtime": 177.0282,
+      "eval_samples_per_second": 56.488,
+      "eval_steps_per_second": 3.531,
+      "step": 93000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7050912584053795e-05,
+      "loss": 2.8096,
+      "step": 93100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.704771053474224e-05,
+      "loss": 2.8016,
+      "step": 93200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7044508485430674e-05,
+      "loss": 2.8249,
+      "step": 93300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.704130643611912e-05,
+      "loss": 2.8016,
+      "step": 93400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.703810438680756e-05,
+      "loss": 2.7854,
+      "step": 93500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7034902337496e-05,
+      "loss": 2.8012,
+      "step": 93600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.703170028818444e-05,
+      "loss": 2.7958,
+      "step": 93700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.702849823887288e-05,
+      "loss": 2.7949,
+      "step": 93800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7025296189561326e-05,
+      "loss": 2.7713,
+      "step": 93900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.702209414024976e-05,
+      "loss": 2.8173,
+      "step": 94000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.594791889190674,
+      "eval_runtime": 177.325,
+      "eval_samples_per_second": 56.394,
+      "eval_steps_per_second": 3.525,
+      "step": 94000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7018892090938205e-05,
+      "loss": 2.7776,
+      "step": 94100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7015690041626645e-05,
+      "loss": 2.8078,
+      "step": 94200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7012487992315085e-05,
+      "loss": 2.7899,
+      "step": 94300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7009285943003525e-05,
+      "loss": 2.7685,
+      "step": 94400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7006083893691964e-05,
+      "loss": 2.7962,
+      "step": 94500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7002881844380404e-05,
+      "loss": 2.7968,
+      "step": 94600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6999679795068844e-05,
+      "loss": 2.7893,
+      "step": 94700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.699647774575729e-05,
+      "loss": 2.7776,
+      "step": 94800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6993275696445723e-05,
+      "loss": 2.8141,
+      "step": 94900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.699007364713417e-05,
+      "loss": 2.7897,
+      "step": 95000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5926544666290283,
+      "eval_runtime": 178.4679,
+      "eval_samples_per_second": 56.033,
+      "eval_steps_per_second": 3.502,
+      "step": 95000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.698687159782261e-05,
+      "loss": 2.8053,
+      "step": 95100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.698366954851105e-05,
+      "loss": 2.7968,
+      "step": 95200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.698046749919949e-05,
+      "loss": 2.7947,
+      "step": 95300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.697726544988793e-05,
+      "loss": 2.7685,
+      "step": 95400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6974063400576375e-05,
+      "loss": 2.8079,
+      "step": 95500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.697086135126481e-05,
+      "loss": 2.7833,
+      "step": 95600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6967659301953255e-05,
+      "loss": 2.7639,
+      "step": 95700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6964457252641695e-05,
+      "loss": 2.7842,
+      "step": 95800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6961255203330134e-05,
+      "loss": 2.7973,
+      "step": 95900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6958053154018574e-05,
+      "loss": 2.79,
+      "step": 96000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.592930555343628,
+      "eval_runtime": 178.0121,
+      "eval_samples_per_second": 56.176,
+      "eval_steps_per_second": 3.511,
+      "step": 96000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6954851104707014e-05,
+      "loss": 2.7893,
+      "step": 96100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.695164905539546e-05,
+      "loss": 2.7867,
+      "step": 96200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.694844700608389e-05,
+      "loss": 2.7888,
+      "step": 96300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.694524495677234e-05,
+      "loss": 2.7878,
+      "step": 96400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.694204290746077e-05,
+      "loss": 2.7961,
+      "step": 96500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.693884085814922e-05,
+      "loss": 2.7752,
+      "step": 96600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.693563880883766e-05,
+      "loss": 2.7838,
+      "step": 96700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.69324367595261e-05,
+      "loss": 2.7856,
+      "step": 96800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.692923471021454e-05,
+      "loss": 2.7888,
+      "step": 96900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.692603266090298e-05,
+      "loss": 2.7849,
+      "step": 97000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.591611623764038,
+      "eval_runtime": 178.4835,
+      "eval_samples_per_second": 56.028,
+      "eval_steps_per_second": 3.502,
+      "step": 97000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6922830611591425e-05,
+      "loss": 2.7922,
+      "step": 97100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.691962856227986e-05,
+      "loss": 2.782,
+      "step": 97200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6916426512968304e-05,
+      "loss": 2.792,
+      "step": 97300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6913224463656744e-05,
+      "loss": 2.7722,
+      "step": 97400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6910022414345184e-05,
+      "loss": 2.7712,
+      "step": 97500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6906820365033623e-05,
+      "loss": 2.7903,
+      "step": 97600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.690361831572206e-05,
+      "loss": 2.7668,
+      "step": 97700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.690041626641051e-05,
+      "loss": 2.8183,
+      "step": 97800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.689721421709894e-05,
+      "loss": 2.7972,
+      "step": 97900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.689401216778739e-05,
+      "loss": 2.7841,
+      "step": 98000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5918164253234863,
+      "eval_runtime": 177.2044,
+      "eval_samples_per_second": 56.432,
+      "eval_steps_per_second": 3.527,
+      "step": 98000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.689081011847582e-05,
+      "loss": 2.7741,
+      "step": 98100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.688760806916427e-05,
+      "loss": 2.7918,
+      "step": 98200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.688440601985271e-05,
+      "loss": 2.7682,
+      "step": 98300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.688120397054115e-05,
+      "loss": 2.7929,
+      "step": 98400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6878001921229595e-05,
+      "loss": 2.7776,
+      "step": 98500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.687479987191803e-05,
+      "loss": 2.799,
+      "step": 98600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6871597822606474e-05,
+      "loss": 2.7845,
+      "step": 98700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.686839577329491e-05,
+      "loss": 2.7769,
+      "step": 98800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6865193723983354e-05,
+      "loss": 2.7855,
+      "step": 98900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.686199167467179e-05,
+      "loss": 2.7917,
+      "step": 99000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5912928581237793,
+      "eval_runtime": 175.7874,
+      "eval_samples_per_second": 56.887,
+      "eval_steps_per_second": 3.555,
+      "step": 99000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.685878962536023e-05,
+      "loss": 2.7712,
+      "step": 99100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.685558757604867e-05,
+      "loss": 2.7852,
+      "step": 99200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.685238552673711e-05,
+      "loss": 2.7987,
+      "step": 99300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.684918347742556e-05,
+      "loss": 2.7848,
+      "step": 99400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.684598142811399e-05,
+      "loss": 2.8,
+      "step": 99500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.684277937880244e-05,
+      "loss": 2.7905,
+      "step": 99600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.683957732949087e-05,
+      "loss": 2.7905,
+      "step": 99700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.683637528017932e-05,
+      "loss": 2.7818,
+      "step": 99800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.683317323086776e-05,
+      "loss": 2.7893,
+      "step": 99900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.68299711815562e-05,
+      "loss": 2.7955,
+      "step": 100000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5897228717803955,
+      "eval_runtime": 175.0685,
+      "eval_samples_per_second": 57.12,
+      "eval_steps_per_second": 3.57,
+      "step": 100000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6826769132244644e-05,
+      "loss": 2.7748,
+      "step": 100100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.682356708293308e-05,
+      "loss": 2.7957,
+      "step": 100200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6820365033621523e-05,
+      "loss": 2.7879,
+      "step": 100300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.6817162984309956e-05,
+      "loss": 2.7955,
+      "step": 100400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.68139609349984e-05,
+      "loss": 2.7963,
+      "step": 100500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.681075888568684e-05,
+      "loss": 2.7765,
+      "step": 100600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.680755683637528e-05,
+      "loss": 2.7767,
+      "step": 100700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.680435478706372e-05,
+      "loss": 2.7808,
+      "step": 100800
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.680115273775216e-05,
+      "loss": 2.7726,
+      "step": 100900
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.679795068844061e-05,
+      "loss": 2.7805,
+      "step": 101000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.5878331661224365,
+      "eval_runtime": 175.8575,
+      "eval_samples_per_second": 56.864,
+      "eval_steps_per_second": 3.554,
+      "step": 101000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.679474863912904e-05,
+      "loss": 2.7823,
+      "step": 101100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.679154658981749e-05,
+      "loss": 2.7645,
+      "step": 101200
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.678834454050592e-05,
+      "loss": 2.7889,
+      "step": 101300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.678514249119437e-05,
+      "loss": 2.7796,
+      "step": 101400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.678194044188281e-05,
+      "loss": 2.7939,
+      "step": 101500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.677873839257125e-05,
+      "loss": 2.7863,
+      "step": 101600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6775536343259693e-05,
+      "loss": 2.7841,
+      "step": 101700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6772334293948126e-05,
+      "loss": 2.7648,
+      "step": 101800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.676913224463657e-05,
+      "loss": 2.7965,
+      "step": 101900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6765930195325006e-05,
+      "loss": 2.7859,
+      "step": 102000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5862410068511963,
+      "eval_runtime": 175.807,
+      "eval_samples_per_second": 56.881,
+      "eval_steps_per_second": 3.555,
+      "step": 102000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.676272814601345e-05,
+      "loss": 2.807,
+      "step": 102100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.675952609670189e-05,
+      "loss": 2.7853,
+      "step": 102200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.675632404739033e-05,
+      "loss": 2.778,
+      "step": 102300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.675312199807877e-05,
+      "loss": 2.7789,
+      "step": 102400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.674991994876721e-05,
+      "loss": 2.7656,
+      "step": 102500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.674671789945566e-05,
+      "loss": 2.7784,
+      "step": 102600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.674351585014409e-05,
+      "loss": 2.7832,
+      "step": 102700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.674031380083254e-05,
+      "loss": 2.779,
+      "step": 102800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.673711175152097e-05,
+      "loss": 2.7935,
+      "step": 102900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.673390970220942e-05,
+      "loss": 2.7702,
+      "step": 103000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5846774578094482,
+      "eval_runtime": 175.2999,
+      "eval_samples_per_second": 57.045,
+      "eval_steps_per_second": 3.565,
+      "step": 103000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6730707652897857e-05,
+      "loss": 2.7767,
+      "step": 103100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6727505603586296e-05,
+      "loss": 2.7939,
+      "step": 103200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.672430355427474e-05,
+      "loss": 2.7715,
+      "step": 103300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6721101504963176e-05,
+      "loss": 2.7721,
+      "step": 103400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.671789945565162e-05,
+      "loss": 2.753,
+      "step": 103500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6714697406340055e-05,
+      "loss": 2.7613,
+      "step": 103600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.67114953570285e-05,
+      "loss": 2.7891,
+      "step": 103700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.670829330771694e-05,
+      "loss": 2.761,
+      "step": 103800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.670509125840538e-05,
+      "loss": 2.7755,
+      "step": 103900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.670188920909382e-05,
+      "loss": 2.7838,
+      "step": 104000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.584876537322998,
+      "eval_runtime": 175.7481,
+      "eval_samples_per_second": 56.9,
+      "eval_steps_per_second": 3.556,
+      "step": 104000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.669868715978226e-05,
+      "loss": 2.7767,
+      "step": 104100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.669548511047071e-05,
+      "loss": 2.7673,
+      "step": 104200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.669228306115914e-05,
+      "loss": 2.7829,
+      "step": 104300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.668908101184759e-05,
+      "loss": 2.7594,
+      "step": 104400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.668587896253602e-05,
+      "loss": 2.7741,
+      "step": 104500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6682676913224466e-05,
+      "loss": 2.7697,
+      "step": 104600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6679474863912906e-05,
+      "loss": 2.782,
+      "step": 104700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6676272814601346e-05,
+      "loss": 2.7872,
+      "step": 104800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.667307076528979e-05,
+      "loss": 2.7659,
+      "step": 104900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6669868715978225e-05,
+      "loss": 2.7707,
+      "step": 105000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5838239192962646,
+      "eval_runtime": 179.8684,
+      "eval_samples_per_second": 55.596,
+      "eval_steps_per_second": 3.475,
+      "step": 105000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 2.7765,
+      "step": 105100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6663464617355105e-05,
+      "loss": 2.7746,
+      "step": 105200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.666026256804355e-05,
+      "loss": 2.7765,
+      "step": 105300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.665706051873199e-05,
+      "loss": 2.7768,
+      "step": 105400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.665385846942043e-05,
+      "loss": 2.7716,
+      "step": 105500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.665065642010887e-05,
+      "loss": 2.7677,
+      "step": 105600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.664745437079731e-05,
+      "loss": 2.7885,
+      "step": 105700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6644252321485757e-05,
+      "loss": 2.7738,
+      "step": 105800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.664105027217419e-05,
+      "loss": 2.7773,
+      "step": 105900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6637848222862636e-05,
+      "loss": 2.7775,
+      "step": 106000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5846199989318848,
+      "eval_runtime": 175.7932,
+      "eval_samples_per_second": 56.885,
+      "eval_steps_per_second": 3.555,
+      "step": 106000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6634646173551076e-05,
+      "loss": 2.7822,
+      "step": 106100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6631444124239516e-05,
+      "loss": 2.7706,
+      "step": 106200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6628242074927955e-05,
+      "loss": 2.7801,
+      "step": 106300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6625040025616395e-05,
+      "loss": 2.7994,
+      "step": 106400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.662183797630484e-05,
+      "loss": 2.7621,
+      "step": 106500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6618635926993274e-05,
+      "loss": 2.7701,
+      "step": 106600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.661543387768172e-05,
+      "loss": 2.7529,
+      "step": 106700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6612231828370154e-05,
+      "loss": 2.7872,
+      "step": 106800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.66090297790586e-05,
+      "loss": 2.7569,
+      "step": 106900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.660582772974704e-05,
+      "loss": 2.7841,
+      "step": 107000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5815844535827637,
+      "eval_runtime": 174.957,
+      "eval_samples_per_second": 57.157,
+      "eval_steps_per_second": 3.572,
+      "step": 107000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.660262568043548e-05,
+      "loss": 2.7862,
+      "step": 107100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.659942363112392e-05,
+      "loss": 2.7593,
+      "step": 107200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.659622158181236e-05,
+      "loss": 2.7682,
+      "step": 107300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6593019532500806e-05,
+      "loss": 2.7627,
+      "step": 107400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.658981748318924e-05,
+      "loss": 2.7803,
+      "step": 107500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6586615433877685e-05,
+      "loss": 2.7773,
+      "step": 107600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6583413384566125e-05,
+      "loss": 2.7752,
+      "step": 107700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6580211335254565e-05,
+      "loss": 2.7701,
+      "step": 107800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6577009285943005e-05,
+      "loss": 2.7795,
+      "step": 107900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6573807236631444e-05,
+      "loss": 2.7827,
+      "step": 108000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.580613374710083,
+      "eval_runtime": 175.685,
+      "eval_samples_per_second": 56.92,
+      "eval_steps_per_second": 3.558,
+      "step": 108000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.657060518731989e-05,
+      "loss": 2.766,
+      "step": 108100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6567403138008324e-05,
+      "loss": 2.7593,
+      "step": 108200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.656420108869677e-05,
+      "loss": 2.7864,
+      "step": 108300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.656099903938521e-05,
+      "loss": 2.8006,
+      "step": 108400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.655779699007365e-05,
+      "loss": 2.7565,
+      "step": 108500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.655459494076209e-05,
+      "loss": 2.787,
+      "step": 108600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.655139289145053e-05,
+      "loss": 2.7661,
+      "step": 108700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.654819084213897e-05,
+      "loss": 2.7825,
+      "step": 108800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.654498879282741e-05,
+      "loss": 2.7784,
+      "step": 108900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6541786743515855e-05,
+      "loss": 2.7816,
+      "step": 109000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5822932720184326,
+      "eval_runtime": 176.0692,
+      "eval_samples_per_second": 56.796,
+      "eval_steps_per_second": 3.55,
+      "step": 109000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.653858469420429e-05,
+      "loss": 2.7873,
+      "step": 109100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6535382644892735e-05,
+      "loss": 2.763,
+      "step": 109200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6532180595581175e-05,
+      "loss": 2.776,
+      "step": 109300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6528978546269614e-05,
+      "loss": 2.7801,
+      "step": 109400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6525776496958054e-05,
+      "loss": 2.7916,
+      "step": 109500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6522574447646494e-05,
+      "loss": 2.7791,
+      "step": 109600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.651937239833494e-05,
+      "loss": 2.7781,
+      "step": 109700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.651617034902337e-05,
+      "loss": 2.7731,
+      "step": 109800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.651296829971182e-05,
+      "loss": 2.7709,
+      "step": 109900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.650976625040026e-05,
+      "loss": 2.7724,
+      "step": 110000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.581916332244873,
+      "eval_runtime": 175.7459,
+      "eval_samples_per_second": 56.9,
+      "eval_steps_per_second": 3.556,
+      "step": 110000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.65065642010887e-05,
+      "loss": 2.7789,
+      "step": 110100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.650336215177714e-05,
+      "loss": 2.7736,
+      "step": 110200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.650016010246558e-05,
+      "loss": 2.7725,
+      "step": 110300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.649695805315402e-05,
+      "loss": 2.7735,
+      "step": 110400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.649375600384246e-05,
+      "loss": 2.7679,
+      "step": 110500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6490553954530905e-05,
+      "loss": 2.7728,
+      "step": 110600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6487351905219344e-05,
+      "loss": 2.7534,
+      "step": 110700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6484149855907784e-05,
+      "loss": 2.773,
+      "step": 110800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6480947806596224e-05,
+      "loss": 2.7793,
+      "step": 110900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6477745757284664e-05,
+      "loss": 2.7658,
+      "step": 111000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5806679725646973,
+      "eval_runtime": 176.035,
+      "eval_samples_per_second": 56.807,
+      "eval_steps_per_second": 3.55,
+      "step": 111000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.64745437079731e-05,
+      "loss": 2.779,
+      "step": 111100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.647134165866154e-05,
+      "loss": 2.7775,
+      "step": 111200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.646813960934999e-05,
+      "loss": 2.7601,
+      "step": 111300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.646493756003843e-05,
+      "loss": 2.7548,
+      "step": 111400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.646173551072687e-05,
+      "loss": 2.7505,
+      "step": 111500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.645853346141531e-05,
+      "loss": 2.7647,
+      "step": 111600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.645533141210375e-05,
+      "loss": 2.7719,
+      "step": 111700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.645212936279219e-05,
+      "loss": 2.7607,
+      "step": 111800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.644892731348063e-05,
+      "loss": 2.7816,
+      "step": 111900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.644572526416907e-05,
+      "loss": 2.7741,
+      "step": 112000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.578455924987793,
+      "eval_runtime": 176.0528,
+      "eval_samples_per_second": 56.801,
+      "eval_steps_per_second": 3.55,
+      "step": 112000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.644252321485751e-05,
+      "loss": 2.7481,
+      "step": 112100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6439321165545954e-05,
+      "loss": 2.751,
+      "step": 112200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6436119116234394e-05,
+      "loss": 2.7685,
+      "step": 112300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6432917066922834e-05,
+      "loss": 2.7622,
+      "step": 112400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.642971501761127e-05,
+      "loss": 2.7674,
+      "step": 112500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.642651296829971e-05,
+      "loss": 2.7388,
+      "step": 112600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.642331091898815e-05,
+      "loss": 2.7537,
+      "step": 112700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.642010886967659e-05,
+      "loss": 2.757,
+      "step": 112800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.641690682036504e-05,
+      "loss": 2.7497,
+      "step": 112900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.641370477105348e-05,
+      "loss": 2.8012,
+      "step": 113000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.575178384780884,
+      "eval_runtime": 176.1143,
+      "eval_samples_per_second": 56.781,
+      "eval_steps_per_second": 3.549,
+      "step": 113000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.641050272174192e-05,
+      "loss": 2.772,
+      "step": 113100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.640730067243036e-05,
+      "loss": 2.7572,
+      "step": 113200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.64040986231188e-05,
+      "loss": 2.763,
+      "step": 113300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.640089657380724e-05,
+      "loss": 2.74,
+      "step": 113400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.639769452449568e-05,
+      "loss": 2.7686,
+      "step": 113500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.639449247518412e-05,
+      "loss": 2.7636,
+      "step": 113600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6391290425872564e-05,
+      "loss": 2.7625,
+      "step": 113700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6388088376561003e-05,
+      "loss": 2.7554,
+      "step": 113800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.638488632724944e-05,
+      "loss": 2.7588,
+      "step": 113900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.638168427793788e-05,
+      "loss": 2.7621,
+      "step": 114000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5783112049102783,
+      "eval_runtime": 178.5692,
+      "eval_samples_per_second": 56.001,
+      "eval_steps_per_second": 3.5,
+      "step": 114000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.637848222862632e-05,
+      "loss": 2.7509,
+      "step": 114100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.637528017931476e-05,
+      "loss": 2.7537,
+      "step": 114200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.63720781300032e-05,
+      "loss": 2.7738,
+      "step": 114300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.636887608069164e-05,
+      "loss": 2.7833,
+      "step": 114400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.636567403138009e-05,
+      "loss": 2.7972,
+      "step": 114500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.636247198206853e-05,
+      "loss": 2.7572,
+      "step": 114600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.635926993275697e-05,
+      "loss": 2.7588,
+      "step": 114700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.635606788344541e-05,
+      "loss": 2.7773,
+      "step": 114800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.635286583413385e-05,
+      "loss": 2.7797,
+      "step": 114900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.634966378482229e-05,
+      "loss": 2.7656,
+      "step": 115000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.577411413192749,
+      "eval_runtime": 175.7584,
+      "eval_samples_per_second": 56.896,
+      "eval_steps_per_second": 3.556,
+      "step": 115000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.634646173551073e-05,
+      "loss": 2.7533,
+      "step": 115100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6343259686199167e-05,
+      "loss": 2.7707,
+      "step": 115200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.634005763688761e-05,
+      "loss": 2.7447,
+      "step": 115300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.633685558757605e-05,
+      "loss": 2.7532,
+      "step": 115400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.633365353826449e-05,
+      "loss": 2.7768,
+      "step": 115500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.633045148895293e-05,
+      "loss": 2.7717,
+      "step": 115600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.632724943964137e-05,
+      "loss": 2.7603,
+      "step": 115700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.632404739032981e-05,
+      "loss": 2.7537,
+      "step": 115800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.632084534101825e-05,
+      "loss": 2.7497,
+      "step": 115900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.63176432917067e-05,
+      "loss": 2.7698,
+      "step": 116000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5763938426971436,
+      "eval_runtime": 176.1475,
+      "eval_samples_per_second": 56.771,
+      "eval_steps_per_second": 3.548,
+      "step": 116000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.631444124239514e-05,
+      "loss": 2.7675,
+      "step": 116100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.631123919308358e-05,
+      "loss": 2.7518,
+      "step": 116200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.630803714377202e-05,
+      "loss": 2.7729,
+      "step": 116300
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.630483509446046e-05,
+      "loss": 2.7543,
+      "step": 116400
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.63016330451489e-05,
+      "loss": 2.752,
+      "step": 116500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6298430995837336e-05,
+      "loss": 2.7687,
+      "step": 116600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6295228946525776e-05,
+      "loss": 2.7495,
+      "step": 116700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.6292026897214216e-05,
+      "loss": 2.7704,
+      "step": 116800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.628882484790266e-05,
+      "loss": 2.7339,
+      "step": 116900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.62856227985911e-05,
+      "loss": 2.7739,
+      "step": 117000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 2.5727386474609375,
+      "eval_runtime": 177.1346,
+      "eval_samples_per_second": 56.454,
+      "eval_steps_per_second": 3.528,
+      "step": 117000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.628242074927954e-05,
+      "loss": 2.7556,
+      "step": 117100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.627921869996798e-05,
+      "loss": 2.7758,
+      "step": 117200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.627601665065642e-05,
+      "loss": 2.7593,
+      "step": 117300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.627281460134486e-05,
+      "loss": 2.7444,
+      "step": 117400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.62696125520333e-05,
+      "loss": 2.7513,
+      "step": 117500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.626641050272175e-05,
+      "loss": 2.7554,
+      "step": 117600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.626320845341019e-05,
+      "loss": 2.7477,
+      "step": 117700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.626000640409863e-05,
+      "loss": 2.7488,
+      "step": 117800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6256804354787067e-05,
+      "loss": 2.7491,
+      "step": 117900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6253602305475506e-05,
+      "loss": 2.7655,
+      "step": 118000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.572309732437134,
+      "eval_runtime": 174.965,
+      "eval_samples_per_second": 57.154,
+      "eval_steps_per_second": 3.572,
+      "step": 118000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6250400256163946e-05,
+      "loss": 2.7691,
+      "step": 118100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6247198206852386e-05,
+      "loss": 2.7688,
+      "step": 118200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.624399615754083e-05,
+      "loss": 2.7428,
+      "step": 118300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6240794108229265e-05,
+      "loss": 2.7508,
+      "step": 118400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.623759205891771e-05,
+      "loss": 2.7109,
+      "step": 118500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.623439000960615e-05,
+      "loss": 2.7504,
+      "step": 118600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.623118796029459e-05,
+      "loss": 2.759,
+      "step": 118700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.622798591098303e-05,
+      "loss": 2.7441,
+      "step": 118800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.622478386167147e-05,
+      "loss": 2.7396,
+      "step": 118900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.622158181235991e-05,
+      "loss": 2.747,
+      "step": 119000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.574122190475464,
+      "eval_runtime": 175.7122,
+      "eval_samples_per_second": 56.911,
+      "eval_steps_per_second": 3.557,
+      "step": 119000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.621837976304835e-05,
+      "loss": 2.757,
+      "step": 119100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.62151777137368e-05,
+      "loss": 2.7434,
+      "step": 119200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6211975664425236e-05,
+      "loss": 2.7499,
+      "step": 119300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6208773615113676e-05,
+      "loss": 2.7546,
+      "step": 119400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6205571565802116e-05,
+      "loss": 2.7485,
+      "step": 119500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6202369516490556e-05,
+      "loss": 2.7285,
+      "step": 119600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6199167467178995e-05,
+      "loss": 2.7476,
+      "step": 119700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6195965417867435e-05,
+      "loss": 2.7547,
+      "step": 119800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.619276336855588e-05,
+      "loss": 2.7497,
+      "step": 119900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6189561319244315e-05,
+      "loss": 2.7221,
+      "step": 120000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.573463201522827,
+      "eval_runtime": 177.4563,
+      "eval_samples_per_second": 56.352,
+      "eval_steps_per_second": 3.522,
+      "step": 120000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.618635926993276e-05,
+      "loss": 2.7397,
+      "step": 120100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.61831572206212e-05,
+      "loss": 2.7323,
+      "step": 120200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.617995517130964e-05,
+      "loss": 2.7262,
+      "step": 120300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.617675312199808e-05,
+      "loss": 2.7333,
+      "step": 120400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.617355107268652e-05,
+      "loss": 2.7472,
+      "step": 120500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.617034902337497e-05,
+      "loss": 2.7451,
+      "step": 120600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.61671469740634e-05,
+      "loss": 2.7302,
+      "step": 120700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6163944924751846e-05,
+      "loss": 2.7399,
+      "step": 120800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6160742875440286e-05,
+      "loss": 2.7264,
+      "step": 120900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6157540826128726e-05,
+      "loss": 2.7283,
+      "step": 121000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5729925632476807,
+      "eval_runtime": 177.6122,
+      "eval_samples_per_second": 56.302,
+      "eval_steps_per_second": 3.519,
+      "step": 121000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6154338776817165e-05,
+      "loss": 2.7362,
+      "step": 121100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6151136727505605e-05,
+      "loss": 2.7247,
+      "step": 121200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.614793467819405e-05,
+      "loss": 2.7468,
+      "step": 121300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6144732628882485e-05,
+      "loss": 2.7575,
+      "step": 121400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.614153057957093e-05,
+      "loss": 2.7403,
+      "step": 121500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6138328530259364e-05,
+      "loss": 2.7355,
+      "step": 121600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.613512648094781e-05,
+      "loss": 2.746,
+      "step": 121700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.613192443163625e-05,
+      "loss": 2.7353,
+      "step": 121800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.612872238232469e-05,
+      "loss": 2.7497,
+      "step": 121900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.612552033301313e-05,
+      "loss": 2.7617,
+      "step": 122000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5720784664154053,
+      "eval_runtime": 176.4862,
+      "eval_samples_per_second": 56.662,
+      "eval_steps_per_second": 3.541,
+      "step": 122000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.612231828370157e-05,
+      "loss": 2.7553,
+      "step": 122100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6119116234390016e-05,
+      "loss": 2.7597,
+      "step": 122200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.611591418507845e-05,
+      "loss": 2.7637,
+      "step": 122300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6112712135766895e-05,
+      "loss": 2.7447,
+      "step": 122400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6109510086455335e-05,
+      "loss": 2.7318,
+      "step": 122500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6106308037143775e-05,
+      "loss": 2.7382,
+      "step": 122600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6103105987832215e-05,
+      "loss": 2.7487,
+      "step": 122700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6099903938520654e-05,
+      "loss": 2.7538,
+      "step": 122800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.60967018892091e-05,
+      "loss": 2.7384,
+      "step": 122900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6093499839897534e-05,
+      "loss": 2.7676,
+      "step": 123000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5681958198547363,
+      "eval_runtime": 175.8725,
+      "eval_samples_per_second": 56.859,
+      "eval_steps_per_second": 3.554,
+      "step": 123000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.609029779058598e-05,
+      "loss": 2.7297,
+      "step": 123100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6087095741274413e-05,
+      "loss": 2.7523,
+      "step": 123200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.608389369196286e-05,
+      "loss": 2.7259,
+      "step": 123300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.60806916426513e-05,
+      "loss": 2.7461,
+      "step": 123400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.607748959333974e-05,
+      "loss": 2.7475,
+      "step": 123500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6074287544028186e-05,
+      "loss": 2.7419,
+      "step": 123600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.607108549471662e-05,
+      "loss": 2.7554,
+      "step": 123700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6067883445405065e-05,
+      "loss": 2.7616,
+      "step": 123800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.60646813960935e-05,
+      "loss": 2.7714,
+      "step": 123900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6061479346781945e-05,
+      "loss": 2.7425,
+      "step": 124000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.56851863861084,
+      "eval_runtime": 175.5851,
+      "eval_samples_per_second": 56.952,
+      "eval_steps_per_second": 3.56,
+      "step": 124000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6058277297470385e-05,
+      "loss": 2.7394,
+      "step": 124100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6055075248158824e-05,
+      "loss": 2.7452,
+      "step": 124200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6051873198847264e-05,
+      "loss": 2.7259,
+      "step": 124300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6048671149535704e-05,
+      "loss": 2.7442,
+      "step": 124400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.604546910022415e-05,
+      "loss": 2.7356,
+      "step": 124500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.604226705091258e-05,
+      "loss": 2.7473,
+      "step": 124600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.603906500160103e-05,
+      "loss": 2.7423,
+      "step": 124700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.603586295228946e-05,
+      "loss": 2.761,
+      "step": 124800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.603266090297791e-05,
+      "loss": 2.7476,
+      "step": 124900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.602945885366635e-05,
+      "loss": 2.7267,
+      "step": 125000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.567716360092163,
+      "eval_runtime": 176.6756,
+      "eval_samples_per_second": 56.601,
+      "eval_steps_per_second": 3.538,
+      "step": 125000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.602625680435479e-05,
+      "loss": 2.7399,
+      "step": 125100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6023054755043235e-05,
+      "loss": 2.7479,
+      "step": 125200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.601985270573167e-05,
+      "loss": 2.7493,
+      "step": 125300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6016650656420115e-05,
+      "loss": 2.7457,
+      "step": 125400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.601344860710855e-05,
+      "loss": 2.7359,
+      "step": 125500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6010246557796994e-05,
+      "loss": 2.7371,
+      "step": 125600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6007044508485434e-05,
+      "loss": 2.7412,
+      "step": 125700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6003842459173874e-05,
+      "loss": 2.7709,
+      "step": 125800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6000640409862313e-05,
+      "loss": 2.7478,
+      "step": 125900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.599743836055075e-05,
+      "loss": 2.7378,
+      "step": 126000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5691282749176025,
+      "eval_runtime": 178.5443,
+      "eval_samples_per_second": 56.009,
+      "eval_steps_per_second": 3.501,
+      "step": 126000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.59942363112392e-05,
+      "loss": 2.7571,
+      "step": 126100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.599103426192763e-05,
+      "loss": 2.7205,
+      "step": 126200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.598783221261608e-05,
+      "loss": 2.7195,
+      "step": 126300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.598463016330451e-05,
+      "loss": 2.7403,
+      "step": 126400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.598142811399296e-05,
+      "loss": 2.7524,
+      "step": 126500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.59782260646814e-05,
+      "loss": 2.7472,
+      "step": 126600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.597502401536984e-05,
+      "loss": 2.7372,
+      "step": 126700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5971821966058285e-05,
+      "loss": 2.736,
+      "step": 126800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.596861991674672e-05,
+      "loss": 2.7493,
+      "step": 126900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5965417867435164e-05,
+      "loss": 2.7429,
+      "step": 127000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5668647289276123,
+      "eval_runtime": 177.1021,
+      "eval_samples_per_second": 56.465,
+      "eval_steps_per_second": 3.529,
+      "step": 127000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.59622158181236e-05,
+      "loss": 2.7402,
+      "step": 127100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5959013768812044e-05,
+      "loss": 2.747,
+      "step": 127200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.595581171950048e-05,
+      "loss": 2.7192,
+      "step": 127300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.595260967018892e-05,
+      "loss": 2.7243,
+      "step": 127400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.594940762087736e-05,
+      "loss": 2.7299,
+      "step": 127500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.59462055715658e-05,
+      "loss": 2.7256,
+      "step": 127600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.594300352225425e-05,
+      "loss": 2.7402,
+      "step": 127700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.593980147294268e-05,
+      "loss": 2.7284,
+      "step": 127800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.593659942363113e-05,
+      "loss": 2.7088,
+      "step": 127900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.593339737431956e-05,
+      "loss": 2.7382,
+      "step": 128000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5666918754577637,
+      "eval_runtime": 177.4538,
+      "eval_samples_per_second": 56.353,
+      "eval_steps_per_second": 3.522,
+      "step": 128000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.593019532500801e-05,
+      "loss": 2.7279,
+      "step": 128100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.592699327569645e-05,
+      "loss": 2.7276,
+      "step": 128200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.592379122638489e-05,
+      "loss": 2.7334,
+      "step": 128300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5920589177073334e-05,
+      "loss": 2.7367,
+      "step": 128400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.591738712776177e-05,
+      "loss": 2.7565,
+      "step": 128500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5914185078450213e-05,
+      "loss": 2.7269,
+      "step": 128600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5910983029138646e-05,
+      "loss": 2.7653,
+      "step": 128700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.590778097982709e-05,
+      "loss": 2.7261,
+      "step": 128800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.590457893051553e-05,
+      "loss": 2.7526,
+      "step": 128900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.590137688120397e-05,
+      "loss": 2.7317,
+      "step": 129000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5652542114257812,
+      "eval_runtime": 177.1478,
+      "eval_samples_per_second": 56.45,
+      "eval_steps_per_second": 3.528,
+      "step": 129000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.589817483189241e-05,
+      "loss": 2.7361,
+      "step": 129100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.589497278258085e-05,
+      "loss": 2.7237,
+      "step": 129200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.58917707332693e-05,
+      "loss": 2.7527,
+      "step": 129300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.588856868395773e-05,
+      "loss": 2.7399,
+      "step": 129400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.588536663464618e-05,
+      "loss": 2.7329,
+      "step": 129500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.588216458533461e-05,
+      "loss": 2.7414,
+      "step": 129600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.587896253602306e-05,
+      "loss": 2.7156,
+      "step": 129700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.58757604867115e-05,
+      "loss": 2.7384,
+      "step": 129800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.587255843739994e-05,
+      "loss": 2.7398,
+      "step": 129900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.586935638808838e-05,
+      "loss": 2.7147,
+      "step": 130000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5645546913146973,
+      "eval_runtime": 178.8022,
+      "eval_samples_per_second": 55.928,
+      "eval_steps_per_second": 3.495,
+      "step": 130000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5866154338776816e-05,
+      "loss": 2.7349,
+      "step": 130100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.586295228946526e-05,
+      "loss": 2.7412,
+      "step": 130200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5859750240153696e-05,
+      "loss": 2.7231,
+      "step": 130300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.585654819084214e-05,
+      "loss": 2.7387,
+      "step": 130400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.585334614153058e-05,
+      "loss": 2.717,
+      "step": 130500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.585014409221902e-05,
+      "loss": 2.7248,
+      "step": 130600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.584694204290746e-05,
+      "loss": 2.7428,
+      "step": 130700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.58437399935959e-05,
+      "loss": 2.7421,
+      "step": 130800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.584053794428435e-05,
+      "loss": 2.7298,
+      "step": 130900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.583733589497278e-05,
+      "loss": 2.7347,
+      "step": 131000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5643277168273926,
+      "eval_runtime": 177.5491,
+      "eval_samples_per_second": 56.322,
+      "eval_steps_per_second": 3.52,
+      "step": 131000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.583413384566123e-05,
+      "loss": 2.7193,
+      "step": 131100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.583093179634967e-05,
+      "loss": 2.7488,
+      "step": 131200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.582772974703811e-05,
+      "loss": 2.7346,
+      "step": 131300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5824527697726547e-05,
+      "loss": 2.742,
+      "step": 131400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5821325648414986e-05,
+      "loss": 2.7283,
+      "step": 131500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.581812359910343e-05,
+      "loss": 2.7401,
+      "step": 131600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5814921549791866e-05,
+      "loss": 2.7387,
+      "step": 131700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.581171950048031e-05,
+      "loss": 2.7359,
+      "step": 131800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5808517451168745e-05,
+      "loss": 2.7505,
+      "step": 131900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.580531540185719e-05,
+      "loss": 2.7317,
+      "step": 132000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 2.5639023780822754,
+      "eval_runtime": 175.6025,
+      "eval_samples_per_second": 56.947,
+      "eval_steps_per_second": 3.559,
+      "step": 132000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.580211335254563e-05,
+      "loss": 2.7439,
+      "step": 132100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.579891130323407e-05,
+      "loss": 2.7339,
+      "step": 132200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.579570925392251e-05,
+      "loss": 2.7229,
+      "step": 132300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.579250720461095e-05,
+      "loss": 2.7235,
+      "step": 132400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.57893051552994e-05,
+      "loss": 2.7449,
+      "step": 132500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.578610310598783e-05,
+      "loss": 2.7339,
+      "step": 132600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.578290105667628e-05,
+      "loss": 2.7365,
+      "step": 132700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5779699007364716e-05,
+      "loss": 2.7345,
+      "step": 132800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5776496958053156e-05,
+      "loss": 2.7119,
+      "step": 132900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5773294908741596e-05,
+      "loss": 2.7415,
+      "step": 133000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5623250007629395,
+      "eval_runtime": 176.2335,
+      "eval_samples_per_second": 56.743,
+      "eval_steps_per_second": 3.546,
+      "step": 133000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5770092859430036e-05,
+      "loss": 2.7264,
+      "step": 133100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.576689081011848e-05,
+      "loss": 2.7485,
+      "step": 133200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5763688760806915e-05,
+      "loss": 2.7293,
+      "step": 133300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.576048671149536e-05,
+      "loss": 2.7132,
+      "step": 133400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.57572846621838e-05,
+      "loss": 2.7241,
+      "step": 133500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.575408261287224e-05,
+      "loss": 2.7353,
+      "step": 133600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.575088056356068e-05,
+      "loss": 2.7326,
+      "step": 133700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.574767851424912e-05,
+      "loss": 2.7297,
+      "step": 133800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.574447646493756e-05,
+      "loss": 2.7046,
+      "step": 133900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5741274415626e-05,
+      "loss": 2.74,
+      "step": 134000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5623319149017334,
+      "eval_runtime": 177.8784,
+      "eval_samples_per_second": 56.218,
+      "eval_steps_per_second": 3.514,
+      "step": 134000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5738072366314447e-05,
+      "loss": 2.7165,
+      "step": 134100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.573487031700288e-05,
+      "loss": 2.733,
+      "step": 134200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5731668267691326e-05,
+      "loss": 2.7403,
+      "step": 134300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5728466218379766e-05,
+      "loss": 2.7304,
+      "step": 134400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5725264169068206e-05,
+      "loss": 2.7204,
+      "step": 134500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5722062119756645e-05,
+      "loss": 2.7082,
+      "step": 134600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5718860070445085e-05,
+      "loss": 2.7246,
+      "step": 134700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.571565802113353e-05,
+      "loss": 2.7432,
+      "step": 134800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5712455971821964e-05,
+      "loss": 2.727,
+      "step": 134900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.570925392251041e-05,
+      "loss": 2.7105,
+      "step": 135000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5619659423828125,
+      "eval_runtime": 175.7613,
+      "eval_samples_per_second": 56.895,
+      "eval_steps_per_second": 3.556,
+      "step": 135000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.570605187319885e-05,
+      "loss": 2.7047,
+      "step": 135100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.570284982388729e-05,
+      "loss": 2.7222,
+      "step": 135200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.569964777457573e-05,
+      "loss": 2.7088,
+      "step": 135300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.569644572526417e-05,
+      "loss": 2.7243,
+      "step": 135400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.569324367595261e-05,
+      "loss": 2.7281,
+      "step": 135500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.569004162664105e-05,
+      "loss": 2.7179,
+      "step": 135600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5686839577329496e-05,
+      "loss": 2.7315,
+      "step": 135700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5683637528017936e-05,
+      "loss": 2.7263,
+      "step": 135800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5680435478706375e-05,
+      "loss": 2.7381,
+      "step": 135900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5677233429394815e-05,
+      "loss": 2.7224,
+      "step": 136000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5620172023773193,
+      "eval_runtime": 177.9081,
+      "eval_samples_per_second": 56.209,
+      "eval_steps_per_second": 3.513,
+      "step": 136000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5674031380083255e-05,
+      "loss": 2.72,
+      "step": 136100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5670829330771695e-05,
+      "loss": 2.7203,
+      "step": 136200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5667627281460134e-05,
+      "loss": 2.7109,
+      "step": 136300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.566442523214858e-05,
+      "loss": 2.7284,
+      "step": 136400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5661223182837014e-05,
+      "loss": 2.7323,
+      "step": 136500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.565802113352546e-05,
+      "loss": 2.7328,
+      "step": 136600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.56548190842139e-05,
+      "loss": 2.7508,
+      "step": 136700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.565161703490234e-05,
+      "loss": 2.7357,
+      "step": 136800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.564841498559078e-05,
+      "loss": 2.7068,
+      "step": 136900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.564521293627922e-05,
+      "loss": 2.6924,
+      "step": 137000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5612192153930664,
+      "eval_runtime": 176.2584,
+      "eval_samples_per_second": 56.735,
+      "eval_steps_per_second": 3.546,
+      "step": 137000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.564201088696766e-05,
+      "loss": 2.7036,
+      "step": 137100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.56388088376561e-05,
+      "loss": 2.7022,
+      "step": 137200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5635606788344545e-05,
+      "loss": 2.7178,
+      "step": 137300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5632404739032985e-05,
+      "loss": 2.7343,
+      "step": 137400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5629202689721425e-05,
+      "loss": 2.7193,
+      "step": 137500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5626000640409865e-05,
+      "loss": 2.7133,
+      "step": 137600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5622798591098304e-05,
+      "loss": 2.7309,
+      "step": 137700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5619596541786744e-05,
+      "loss": 2.7199,
+      "step": 137800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5616394492475184e-05,
+      "loss": 2.7136,
+      "step": 137900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.561319244316363e-05,
+      "loss": 2.7224,
+      "step": 138000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.562304973602295,
+      "eval_runtime": 177.3201,
+      "eval_samples_per_second": 56.395,
+      "eval_steps_per_second": 3.525,
+      "step": 138000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.560999039385207e-05,
+      "loss": 2.7246,
+      "step": 138100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.560678834454051e-05,
+      "loss": 2.7087,
+      "step": 138200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.560358629522895e-05,
+      "loss": 2.7094,
+      "step": 138300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.560038424591739e-05,
+      "loss": 2.7054,
+      "step": 138400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.559718219660583e-05,
+      "loss": 2.7108,
+      "step": 138500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.559398014729427e-05,
+      "loss": 2.7208,
+      "step": 138600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.559077809798271e-05,
+      "loss": 2.7004,
+      "step": 138700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.558757604867115e-05,
+      "loss": 2.7181,
+      "step": 138800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5584373999359595e-05,
+      "loss": 2.7302,
+      "step": 138900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5581171950048034e-05,
+      "loss": 2.7118,
+      "step": 139000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.559722661972046,
+      "eval_runtime": 175.7826,
+      "eval_samples_per_second": 56.888,
+      "eval_steps_per_second": 3.556,
+      "step": 139000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5577969900736474e-05,
+      "loss": 2.7417,
+      "step": 139100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5574767851424914e-05,
+      "loss": 2.7227,
+      "step": 139200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5571565802113354e-05,
+      "loss": 2.7247,
+      "step": 139300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.556836375280179e-05,
+      "loss": 2.7179,
+      "step": 139400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.556516170349023e-05,
+      "loss": 2.7423,
+      "step": 139500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.556195965417868e-05,
+      "loss": 2.7376,
+      "step": 139600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.555875760486712e-05,
+      "loss": 2.7304,
+      "step": 139700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 2.6881,
+      "step": 139800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5552353506244e-05,
+      "loss": 2.7232,
+      "step": 139900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.554915145693244e-05,
+      "loss": 2.7231,
+      "step": 140000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5592551231384277,
+      "eval_runtime": 176.3317,
+      "eval_samples_per_second": 56.711,
+      "eval_steps_per_second": 3.544,
+      "step": 140000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.554594940762088e-05,
+      "loss": 2.7048,
+      "step": 140100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.554274735830932e-05,
+      "loss": 2.7048,
+      "step": 140200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.553954530899776e-05,
+      "loss": 2.7241,
+      "step": 140300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5536343259686204e-05,
+      "loss": 2.7153,
+      "step": 140400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5533141210374644e-05,
+      "loss": 2.7301,
+      "step": 140500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5529939161063084e-05,
+      "loss": 2.7249,
+      "step": 140600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5526737111751524e-05,
+      "loss": 2.7163,
+      "step": 140700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.552353506243996e-05,
+      "loss": 2.722,
+      "step": 140800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.55203330131284e-05,
+      "loss": 2.7067,
+      "step": 140900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.551713096381684e-05,
+      "loss": 2.7196,
+      "step": 141000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5608913898468018,
+      "eval_runtime": 191.9454,
+      "eval_samples_per_second": 52.098,
+      "eval_steps_per_second": 3.256,
+      "step": 141000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.551392891450529e-05,
+      "loss": 2.7199,
+      "step": 141100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.551072686519373e-05,
+      "loss": 2.7258,
+      "step": 141200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.550752481588217e-05,
+      "loss": 2.718,
+      "step": 141300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.550432276657061e-05,
+      "loss": 2.7175,
+      "step": 141400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.550112071725905e-05,
+      "loss": 2.7386,
+      "step": 141500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.549791866794749e-05,
+      "loss": 2.724,
+      "step": 141600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.549471661863593e-05,
+      "loss": 2.6896,
+      "step": 141700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.549151456932437e-05,
+      "loss": 2.715,
+      "step": 141800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.548831252001281e-05,
+      "loss": 2.6942,
+      "step": 141900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5485110470701254e-05,
+      "loss": 2.7205,
+      "step": 142000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5587964057922363,
+      "eval_runtime": 195.6699,
+      "eval_samples_per_second": 51.106,
+      "eval_steps_per_second": 3.194,
+      "step": 142000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5481908421389693e-05,
+      "loss": 2.7107,
+      "step": 142100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.547870637207813e-05,
+      "loss": 2.7236,
+      "step": 142200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.547550432276657e-05,
+      "loss": 2.7218,
+      "step": 142300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.547230227345501e-05,
+      "loss": 2.7444,
+      "step": 142400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.546910022414345e-05,
+      "loss": 2.7069,
+      "step": 142500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.546589817483189e-05,
+      "loss": 2.7037,
+      "step": 142600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.546269612552034e-05,
+      "loss": 2.7266,
+      "step": 142700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.545949407620878e-05,
+      "loss": 2.7243,
+      "step": 142800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.545629202689722e-05,
+      "loss": 2.726,
+      "step": 142900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.545308997758566e-05,
+      "loss": 2.722,
+      "step": 143000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5569887161254883,
+      "eval_runtime": 195.9697,
+      "eval_samples_per_second": 51.028,
+      "eval_steps_per_second": 3.189,
+      "step": 143000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.54498879282741e-05,
+      "loss": 2.7214,
+      "step": 143100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.544668587896254e-05,
+      "loss": 2.7294,
+      "step": 143200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.544348382965098e-05,
+      "loss": 2.7132,
+      "step": 143300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5440281780339424e-05,
+      "loss": 2.7248,
+      "step": 143400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5437079731027857e-05,
+      "loss": 2.6972,
+      "step": 143500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.54338776817163e-05,
+      "loss": 2.7372,
+      "step": 143600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.543067563240474e-05,
+      "loss": 2.7202,
+      "step": 143700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.542747358309318e-05,
+      "loss": 2.7169,
+      "step": 143800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.542427153378162e-05,
+      "loss": 2.7168,
+      "step": 143900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.542106948447006e-05,
+      "loss": 2.7124,
+      "step": 144000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.557919979095459,
+      "eval_runtime": 196.9517,
+      "eval_samples_per_second": 50.774,
+      "eval_steps_per_second": 3.173,
+      "step": 144000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.54178674351585e-05,
+      "loss": 2.6882,
+      "step": 144100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.541466538584694e-05,
+      "loss": 2.7108,
+      "step": 144200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.541146333653539e-05,
+      "loss": 2.7141,
+      "step": 144300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.540826128722383e-05,
+      "loss": 2.6942,
+      "step": 144400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.540505923791227e-05,
+      "loss": 2.7091,
+      "step": 144500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.540185718860071e-05,
+      "loss": 2.7041,
+      "step": 144600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.539865513928915e-05,
+      "loss": 2.7005,
+      "step": 144700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.539545308997759e-05,
+      "loss": 2.6966,
+      "step": 144800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5392251040666026e-05,
+      "loss": 2.7228,
+      "step": 144900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.538904899135447e-05,
+      "loss": 2.7238,
+      "step": 145000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5575151443481445,
+      "eval_runtime": 192.2068,
+      "eval_samples_per_second": 52.027,
+      "eval_steps_per_second": 3.252,
+      "step": 145000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5385846942042906e-05,
+      "loss": 2.7109,
+      "step": 145100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.538264489273135e-05,
+      "loss": 2.7055,
+      "step": 145200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.537944284341979e-05,
+      "loss": 2.7007,
+      "step": 145300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.537624079410823e-05,
+      "loss": 2.7377,
+      "step": 145400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.537303874479667e-05,
+      "loss": 2.6945,
+      "step": 145500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.536983669548511e-05,
+      "loss": 2.7092,
+      "step": 145600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.536663464617356e-05,
+      "loss": 2.7244,
+      "step": 145700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.536343259686199e-05,
+      "loss": 2.7308,
+      "step": 145800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.536023054755044e-05,
+      "loss": 2.6933,
+      "step": 145900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.535702849823888e-05,
+      "loss": 2.7148,
+      "step": 146000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.555891752243042,
+      "eval_runtime": 176.8889,
+      "eval_samples_per_second": 56.533,
+      "eval_steps_per_second": 3.533,
+      "step": 146000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.535382644892732e-05,
+      "loss": 2.7073,
+      "step": 146100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5350624399615757e-05,
+      "loss": 2.7099,
+      "step": 146200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5347422350304196e-05,
+      "loss": 2.7376,
+      "step": 146300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5344220300992636e-05,
+      "loss": 2.722,
+      "step": 146400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5341018251681076e-05,
+      "loss": 2.7329,
+      "step": 146500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.533781620236952e-05,
+      "loss": 2.7116,
+      "step": 146600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5334614153057955e-05,
+      "loss": 2.7152,
+      "step": 146700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.53314121037464e-05,
+      "loss": 2.7132,
+      "step": 146800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.532821005443484e-05,
+      "loss": 2.7152,
+      "step": 146900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.532500800512328e-05,
+      "loss": 2.7224,
+      "step": 147000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5564117431640625,
+      "eval_runtime": 176.4467,
+      "eval_samples_per_second": 56.674,
+      "eval_steps_per_second": 3.542,
+      "step": 147000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.532180595581172e-05,
+      "loss": 2.7205,
+      "step": 147100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.531860390650016e-05,
+      "loss": 2.6828,
+      "step": 147200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.531540185718861e-05,
+      "loss": 2.7089,
+      "step": 147300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.531219980787704e-05,
+      "loss": 2.7076,
+      "step": 147400
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.530899775856549e-05,
+      "loss": 2.7319,
+      "step": 147500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5305795709253926e-05,
+      "loss": 2.7201,
+      "step": 147600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5302593659942366e-05,
+      "loss": 2.7186,
+      "step": 147700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5299391610630806e-05,
+      "loss": 2.7015,
+      "step": 147800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5296189561319246e-05,
+      "loss": 2.693,
+      "step": 147900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.529298751200769e-05,
+      "loss": 2.7169,
+      "step": 148000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 2.5550851821899414,
+      "eval_runtime": 177.3833,
+      "eval_samples_per_second": 56.375,
+      "eval_steps_per_second": 3.523,
+      "step": 148000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5289785462696125e-05,
+      "loss": 2.7141,
+      "step": 148100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.528658341338457e-05,
+      "loss": 2.7082,
+      "step": 148200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5283381364073005e-05,
+      "loss": 2.6801,
+      "step": 148300
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.528017931476145e-05,
+      "loss": 2.7095,
+      "step": 148400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.527697726544989e-05,
+      "loss": 2.7016,
+      "step": 148500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.527377521613833e-05,
+      "loss": 2.7217,
+      "step": 148600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.527057316682678e-05,
+      "loss": 2.6865,
+      "step": 148700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.526737111751521e-05,
+      "loss": 2.7088,
+      "step": 148800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.526416906820366e-05,
+      "loss": 2.6978,
+      "step": 148900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.526096701889209e-05,
+      "loss": 2.7199,
+      "step": 149000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5551364421844482,
+      "eval_runtime": 176.4025,
+      "eval_samples_per_second": 56.689,
+      "eval_steps_per_second": 3.543,
+      "step": 149000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5257764969580536e-05,
+      "loss": 2.7284,
+      "step": 149100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5254562920268976e-05,
+      "loss": 2.7104,
+      "step": 149200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5251360870957416e-05,
+      "loss": 2.6976,
+      "step": 149300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5248158821645855e-05,
+      "loss": 2.6883,
+      "step": 149400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5244956772334295e-05,
+      "loss": 2.7103,
+      "step": 149500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.524175472302274e-05,
+      "loss": 2.6983,
+      "step": 149600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5238552673711175e-05,
+      "loss": 2.7162,
+      "step": 149700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.523535062439962e-05,
+      "loss": 2.699,
+      "step": 149800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5232148575088054e-05,
+      "loss": 2.7355,
+      "step": 149900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.52289465257765e-05,
+      "loss": 2.696,
+      "step": 150000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5558013916015625,
+      "eval_runtime": 176.7584,
+      "eval_samples_per_second": 56.574,
+      "eval_steps_per_second": 3.536,
+      "step": 150000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.522574447646494e-05,
+      "loss": 2.7101,
+      "step": 150100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.522254242715338e-05,
+      "loss": 2.6998,
+      "step": 150200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5219340377841827e-05,
+      "loss": 2.7,
+      "step": 150300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.521613832853026e-05,
+      "loss": 2.6892,
+      "step": 150400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5212936279218706e-05,
+      "loss": 2.7206,
+      "step": 150500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.520973422990714e-05,
+      "loss": 2.7221,
+      "step": 150600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5206532180595585e-05,
+      "loss": 2.6952,
+      "step": 150700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5203330131284025e-05,
+      "loss": 2.7009,
+      "step": 150800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5200128081972465e-05,
+      "loss": 2.7042,
+      "step": 150900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5196926032660905e-05,
+      "loss": 2.7094,
+      "step": 151000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5532150268554688,
+      "eval_runtime": 176.648,
+      "eval_samples_per_second": 56.61,
+      "eval_steps_per_second": 3.538,
+      "step": 151000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5193723983349344e-05,
+      "loss": 2.6896,
+      "step": 151100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.519052193403779e-05,
+      "loss": 2.7063,
+      "step": 151200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5187319884726224e-05,
+      "loss": 2.728,
+      "step": 151300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.518411783541467e-05,
+      "loss": 2.6979,
+      "step": 151400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.51809157861031e-05,
+      "loss": 2.7126,
+      "step": 151500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.517771373679155e-05,
+      "loss": 2.7166,
+      "step": 151600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.517451168747999e-05,
+      "loss": 2.6942,
+      "step": 151700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.517130963816843e-05,
+      "loss": 2.6964,
+      "step": 151800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5168107588856876e-05,
+      "loss": 2.709,
+      "step": 151900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.516490553954531e-05,
+      "loss": 2.7005,
+      "step": 152000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5514307022094727,
+      "eval_runtime": 176.7201,
+      "eval_samples_per_second": 56.587,
+      "eval_steps_per_second": 3.537,
+      "step": 152000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5161703490233755e-05,
+      "loss": 2.7171,
+      "step": 152100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.515850144092219e-05,
+      "loss": 2.6838,
+      "step": 152200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5155299391610635e-05,
+      "loss": 2.7012,
+      "step": 152300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.515209734229907e-05,
+      "loss": 2.6938,
+      "step": 152400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5148895292987514e-05,
+      "loss": 2.7044,
+      "step": 152500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5145693243675954e-05,
+      "loss": 2.6927,
+      "step": 152600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5142491194364394e-05,
+      "loss": 2.6975,
+      "step": 152700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.513928914505284e-05,
+      "loss": 2.7127,
+      "step": 152800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.513608709574127e-05,
+      "loss": 2.7062,
+      "step": 152900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.513288504642972e-05,
+      "loss": 2.6974,
+      "step": 153000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.552427053451538,
+      "eval_runtime": 176.0172,
+      "eval_samples_per_second": 56.813,
+      "eval_steps_per_second": 3.551,
+      "step": 153000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.512968299711815e-05,
+      "loss": 2.7045,
+      "step": 153100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.51264809478066e-05,
+      "loss": 2.7105,
+      "step": 153200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.512327889849504e-05,
+      "loss": 2.7038,
+      "step": 153300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.512007684918348e-05,
+      "loss": 2.7134,
+      "step": 153400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5116874799871925e-05,
+      "loss": 2.7209,
+      "step": 153500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.511367275056036e-05,
+      "loss": 2.69,
+      "step": 153600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5110470701248805e-05,
+      "loss": 2.7025,
+      "step": 153700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.510726865193724e-05,
+      "loss": 2.6883,
+      "step": 153800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5104066602625684e-05,
+      "loss": 2.7054,
+      "step": 153900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.510086455331412e-05,
+      "loss": 2.7031,
+      "step": 154000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5516436100006104,
+      "eval_runtime": 175.5362,
+      "eval_samples_per_second": 56.968,
+      "eval_steps_per_second": 3.561,
+      "step": 154000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5097662504002564e-05,
+      "loss": 2.7158,
+      "step": 154100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5094460454691003e-05,
+      "loss": 2.7083,
+      "step": 154200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.509125840537944e-05,
+      "loss": 2.6873,
+      "step": 154300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.508805635606789e-05,
+      "loss": 2.7017,
+      "step": 154400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.508485430675632e-05,
+      "loss": 2.7127,
+      "step": 154500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.508165225744477e-05,
+      "loss": 2.6783,
+      "step": 154600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.50784502081332e-05,
+      "loss": 2.698,
+      "step": 154700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.507524815882165e-05,
+      "loss": 2.6893,
+      "step": 154800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.507204610951009e-05,
+      "loss": 2.7094,
+      "step": 154900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.506884406019853e-05,
+      "loss": 2.6675,
+      "step": 155000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5507030487060547,
+      "eval_runtime": 176.7491,
+      "eval_samples_per_second": 56.577,
+      "eval_steps_per_second": 3.536,
+      "step": 155000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5065642010886975e-05,
+      "loss": 2.7306,
+      "step": 155100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.506243996157541e-05,
+      "loss": 2.7059,
+      "step": 155200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5059237912263854e-05,
+      "loss": 2.7102,
+      "step": 155300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.505603586295229e-05,
+      "loss": 2.6915,
+      "step": 155400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5052833813640734e-05,
+      "loss": 2.7161,
+      "step": 155500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.504963176432917e-05,
+      "loss": 2.6955,
+      "step": 155600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.504642971501761e-05,
+      "loss": 2.7184,
+      "step": 155700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.504322766570605e-05,
+      "loss": 2.6794,
+      "step": 155800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.504002561639449e-05,
+      "loss": 2.6808,
+      "step": 155900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.503682356708294e-05,
+      "loss": 2.691,
+      "step": 156000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.550930976867676,
+      "eval_runtime": 176.1204,
+      "eval_samples_per_second": 56.779,
+      "eval_steps_per_second": 3.549,
+      "step": 156000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.503362151777137e-05,
+      "loss": 2.7063,
+      "step": 156100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.503041946845982e-05,
+      "loss": 2.7128,
+      "step": 156200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.502721741914825e-05,
+      "loss": 2.7084,
+      "step": 156300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.50240153698367e-05,
+      "loss": 2.6989,
+      "step": 156400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.502081332052514e-05,
+      "loss": 2.6856,
+      "step": 156500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.501761127121358e-05,
+      "loss": 2.7051,
+      "step": 156600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5014409221902024e-05,
+      "loss": 2.6999,
+      "step": 156700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.501120717259046e-05,
+      "loss": 2.6964,
+      "step": 156800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5008005123278903e-05,
+      "loss": 2.6841,
+      "step": 156900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5004803073967336e-05,
+      "loss": 2.7058,
+      "step": 157000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5504448413848877,
+      "eval_runtime": 176.3728,
+      "eval_samples_per_second": 56.698,
+      "eval_steps_per_second": 3.544,
+      "step": 157000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.500160102465578e-05,
+      "loss": 2.6789,
+      "step": 157100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.499839897534422e-05,
+      "loss": 2.6787,
+      "step": 157200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.499519692603266e-05,
+      "loss": 2.6961,
+      "step": 157300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.49919948767211e-05,
+      "loss": 2.6976,
+      "step": 157400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.498879282740954e-05,
+      "loss": 2.7085,
+      "step": 157500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.498559077809799e-05,
+      "loss": 2.6969,
+      "step": 157600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.498238872878642e-05,
+      "loss": 2.6797,
+      "step": 157700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.497918667947487e-05,
+      "loss": 2.694,
+      "step": 157800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.497598463016331e-05,
+      "loss": 2.6965,
+      "step": 157900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.497278258085175e-05,
+      "loss": 2.6927,
+      "step": 158000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5523300170898438,
+      "eval_runtime": 176.4117,
+      "eval_samples_per_second": 56.686,
+      "eval_steps_per_second": 3.543,
+      "step": 158000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.496958053154019e-05,
+      "loss": 2.6993,
+      "step": 158100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.496637848222863e-05,
+      "loss": 2.6749,
+      "step": 158200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.496317643291707e-05,
+      "loss": 2.6937,
+      "step": 158300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4959974383605506e-05,
+      "loss": 2.6949,
+      "step": 158400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.495677233429395e-05,
+      "loss": 2.689,
+      "step": 158500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.495357028498239e-05,
+      "loss": 2.6996,
+      "step": 158600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.495036823567083e-05,
+      "loss": 2.6969,
+      "step": 158700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.494716618635927e-05,
+      "loss": 2.7083,
+      "step": 158800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.494396413704771e-05,
+      "loss": 2.7004,
+      "step": 158900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.494076208773615e-05,
+      "loss": 2.705,
+      "step": 159000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5506510734558105,
+      "eval_runtime": 176.2949,
+      "eval_samples_per_second": 56.723,
+      "eval_steps_per_second": 3.545,
+      "step": 159000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.493756003842459e-05,
+      "loss": 2.6878,
+      "step": 159100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.493435798911304e-05,
+      "loss": 2.7245,
+      "step": 159200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.493115593980147e-05,
+      "loss": 2.7018,
+      "step": 159300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.492795389048992e-05,
+      "loss": 2.6853,
+      "step": 159400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.492475184117836e-05,
+      "loss": 2.6791,
+      "step": 159500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.49215497918668e-05,
+      "loss": 2.6969,
+      "step": 159600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4918347742555236e-05,
+      "loss": 2.6739,
+      "step": 159700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4915145693243676e-05,
+      "loss": 2.6825,
+      "step": 159800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.491194364393212e-05,
+      "loss": 2.7052,
+      "step": 159900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4908741594620556e-05,
+      "loss": 2.6776,
+      "step": 160000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.548813581466675,
+      "eval_runtime": 176.2094,
+      "eval_samples_per_second": 56.751,
+      "eval_steps_per_second": 3.547,
+      "step": 160000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4905539545309e-05,
+      "loss": 2.6847,
+      "step": 160100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.490233749599744e-05,
+      "loss": 2.6888,
+      "step": 160200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.489913544668588e-05,
+      "loss": 2.6876,
+      "step": 160300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.489593339737432e-05,
+      "loss": 2.6889,
+      "step": 160400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.489273134806276e-05,
+      "loss": 2.6921,
+      "step": 160500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.48895292987512e-05,
+      "loss": 2.706,
+      "step": 160600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.488632724943964e-05,
+      "loss": 2.7059,
+      "step": 160700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.488312520012809e-05,
+      "loss": 2.7196,
+      "step": 160800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.487992315081653e-05,
+      "loss": 2.723,
+      "step": 160900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.487672110150497e-05,
+      "loss": 2.6698,
+      "step": 161000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5499532222747803,
+      "eval_runtime": 177.0923,
+      "eval_samples_per_second": 56.468,
+      "eval_steps_per_second": 3.529,
+      "step": 161000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4873519052193406e-05,
+      "loss": 2.6885,
+      "step": 161100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4870317002881846e-05,
+      "loss": 2.6578,
+      "step": 161200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4867114953570286e-05,
+      "loss": 2.6916,
+      "step": 161300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4863912904258726e-05,
+      "loss": 2.7043,
+      "step": 161400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.486071085494717e-05,
+      "loss": 2.688,
+      "step": 161500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4857508805635605e-05,
+      "loss": 2.6949,
+      "step": 161600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.485430675632405e-05,
+      "loss": 2.6942,
+      "step": 161700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.485110470701249e-05,
+      "loss": 2.6822,
+      "step": 161800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.484790265770093e-05,
+      "loss": 2.6737,
+      "step": 161900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.484470060838937e-05,
+      "loss": 2.7036,
+      "step": 162000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.547950506210327,
+      "eval_runtime": 175.2664,
+      "eval_samples_per_second": 57.056,
+      "eval_steps_per_second": 3.566,
+      "step": 162000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.484149855907781e-05,
+      "loss": 2.7313,
+      "step": 162100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.483829650976625e-05,
+      "loss": 2.6729,
+      "step": 162200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.483509446045469e-05,
+      "loss": 2.6746,
+      "step": 162300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4831892411143137e-05,
+      "loss": 2.7017,
+      "step": 162400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4828690361831576e-05,
+      "loss": 2.6856,
+      "step": 162500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4825488312520016e-05,
+      "loss": 2.6974,
+      "step": 162600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4822286263208456e-05,
+      "loss": 2.697,
+      "step": 162700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4819084213896896e-05,
+      "loss": 2.6867,
+      "step": 162800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4815882164585335e-05,
+      "loss": 2.6919,
+      "step": 162900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4812680115273775e-05,
+      "loss": 2.6869,
+      "step": 163000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5488507747650146,
+      "eval_runtime": 178.2486,
+      "eval_samples_per_second": 56.101,
+      "eval_steps_per_second": 3.506,
+      "step": 163000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.480947806596222e-05,
+      "loss": 2.6961,
+      "step": 163100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.480627601665066e-05,
+      "loss": 2.6871,
+      "step": 163200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.48030739673391e-05,
+      "loss": 2.6938,
+      "step": 163300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.479987191802754e-05,
+      "loss": 2.6779,
+      "step": 163400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.479666986871598e-05,
+      "loss": 2.6705,
+      "step": 163500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.479346781940442e-05,
+      "loss": 2.6818,
+      "step": 163600
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.479026577009286e-05,
+      "loss": 2.679,
+      "step": 163700
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.47870637207813e-05,
+      "loss": 2.67,
+      "step": 163800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.478386167146974e-05,
+      "loss": 2.6943,
+      "step": 163900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4780659622158186e-05,
+      "loss": 2.6805,
+      "step": 164000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.5475895404815674,
+      "eval_runtime": 178.95,
+      "eval_samples_per_second": 55.882,
+      "eval_steps_per_second": 3.493,
+      "step": 164000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4777457572846626e-05,
+      "loss": 2.6768,
+      "step": 164100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4774255523535065e-05,
+      "loss": 2.6859,
+      "step": 164200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4771053474223505e-05,
+      "loss": 2.6726,
+      "step": 164300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4767851424911945e-05,
+      "loss": 2.6899,
+      "step": 164400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4764649375600385e-05,
+      "loss": 2.6961,
+      "step": 164500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4761447326288824e-05,
+      "loss": 2.6584,
+      "step": 164600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.475824527697727e-05,
+      "loss": 2.6667,
+      "step": 164700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.475504322766571e-05,
+      "loss": 2.6932,
+      "step": 164800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.475184117835415e-05,
+      "loss": 2.6838,
+      "step": 164900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.474863912904259e-05,
+      "loss": 2.6931,
+      "step": 165000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.545959711074829,
+      "eval_runtime": 183.1837,
+      "eval_samples_per_second": 54.59,
+      "eval_steps_per_second": 3.412,
+      "step": 165000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.474543707973103e-05,
+      "loss": 2.6931,
+      "step": 165100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.474223503041947e-05,
+      "loss": 2.6918,
+      "step": 165200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.473903298110791e-05,
+      "loss": 2.6938,
+      "step": 165300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.473583093179635e-05,
+      "loss": 2.7014,
+      "step": 165400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4732628882484796e-05,
+      "loss": 2.7183,
+      "step": 165500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4729426833173235e-05,
+      "loss": 2.6951,
+      "step": 165600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4726224783861675e-05,
+      "loss": 2.6888,
+      "step": 165700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4723022734550115e-05,
+      "loss": 2.6841,
+      "step": 165800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4719820685238555e-05,
+      "loss": 2.6971,
+      "step": 165900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4716618635926994e-05,
+      "loss": 2.6807,
+      "step": 166000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.546175241470337,
+      "eval_runtime": 182.6607,
+      "eval_samples_per_second": 54.746,
+      "eval_steps_per_second": 3.422,
+      "step": 166000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4713416586615434e-05,
+      "loss": 2.69,
+      "step": 166100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4710214537303874e-05,
+      "loss": 2.6925,
+      "step": 166200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.470701248799232e-05,
+      "loss": 2.6786,
+      "step": 166300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.470381043868076e-05,
+      "loss": 2.6916,
+      "step": 166400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.47006083893692e-05,
+      "loss": 2.6914,
+      "step": 166500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.469740634005764e-05,
+      "loss": 2.6732,
+      "step": 166600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.469420429074608e-05,
+      "loss": 2.7,
+      "step": 166700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.469100224143452e-05,
+      "loss": 2.6885,
+      "step": 166800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.468780019212296e-05,
+      "loss": 2.6889,
+      "step": 166900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.46845981428114e-05,
+      "loss": 2.6913,
+      "step": 167000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.544877052307129,
+      "eval_runtime": 181.3638,
+      "eval_samples_per_second": 55.138,
+      "eval_steps_per_second": 3.446,
+      "step": 167000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4681396093499845e-05,
+      "loss": 2.6865,
+      "step": 167100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4678194044188285e-05,
+      "loss": 2.6664,
+      "step": 167200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4674991994876724e-05,
+      "loss": 2.6831,
+      "step": 167300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4671789945565164e-05,
+      "loss": 2.6744,
+      "step": 167400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4668587896253604e-05,
+      "loss": 2.6984,
+      "step": 167500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4665385846942044e-05,
+      "loss": 2.7074,
+      "step": 167600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.466218379763048e-05,
+      "loss": 2.6687,
+      "step": 167700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.465898174831893e-05,
+      "loss": 2.686,
+      "step": 167800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.465577969900737e-05,
+      "loss": 2.6705,
+      "step": 167900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.465257764969581e-05,
+      "loss": 2.6895,
+      "step": 168000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.544931173324585,
+      "eval_runtime": 181.9047,
+      "eval_samples_per_second": 54.974,
+      "eval_steps_per_second": 3.436,
+      "step": 168000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.464937560038425e-05,
+      "loss": 2.6817,
+      "step": 168100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.464617355107269e-05,
+      "loss": 2.6892,
+      "step": 168200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.464297150176113e-05,
+      "loss": 2.6709,
+      "step": 168300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.463976945244957e-05,
+      "loss": 2.662,
+      "step": 168400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4636567403138015e-05,
+      "loss": 2.6897,
+      "step": 168500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.463336535382645e-05,
+      "loss": 2.6811,
+      "step": 168600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4630163304514894e-05,
+      "loss": 2.6794,
+      "step": 168700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4626961255203334e-05,
+      "loss": 2.688,
+      "step": 168800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4623759205891774e-05,
+      "loss": 2.668,
+      "step": 168900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4620557156580214e-05,
+      "loss": 2.6813,
+      "step": 169000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.545745611190796,
+      "eval_runtime": 185.2342,
+      "eval_samples_per_second": 53.986,
+      "eval_steps_per_second": 3.374,
+      "step": 169000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.461735510726865e-05,
+      "loss": 2.6721,
+      "step": 169100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.461415305795709e-05,
+      "loss": 2.6777,
+      "step": 169200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.461095100864553e-05,
+      "loss": 2.6988,
+      "step": 169300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.460774895933398e-05,
+      "loss": 2.6658,
+      "step": 169400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.460454691002241e-05,
+      "loss": 2.6527,
+      "step": 169500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.460134486071086e-05,
+      "loss": 2.6751,
+      "step": 169600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.45981428113993e-05,
+      "loss": 2.6917,
+      "step": 169700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.459494076208774e-05,
+      "loss": 2.6868,
+      "step": 169800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.459173871277618e-05,
+      "loss": 2.683,
+      "step": 169900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.458853666346462e-05,
+      "loss": 2.6483,
+      "step": 170000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.546018600463867,
+      "eval_runtime": 178.1131,
+      "eval_samples_per_second": 56.144,
+      "eval_steps_per_second": 3.509,
+      "step": 170000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4585334614153064e-05,
+      "loss": 2.6927,
+      "step": 170100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.45821325648415e-05,
+      "loss": 2.6673,
+      "step": 170200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4578930515529944e-05,
+      "loss": 2.6984,
+      "step": 170300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4575728466218383e-05,
+      "loss": 2.6714,
+      "step": 170400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.457252641690682e-05,
+      "loss": 2.6531,
+      "step": 170500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.456932436759526e-05,
+      "loss": 2.6632,
+      "step": 170600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.45661223182837e-05,
+      "loss": 2.6671,
+      "step": 170700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.456292026897215e-05,
+      "loss": 2.6924,
+      "step": 170800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.455971821966058e-05,
+      "loss": 2.6658,
+      "step": 170900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.455651617034903e-05,
+      "loss": 2.7026,
+      "step": 171000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.5437819957733154,
+      "eval_runtime": 179.7594,
+      "eval_samples_per_second": 55.63,
+      "eval_steps_per_second": 3.477,
+      "step": 171000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.455331412103746e-05,
+      "loss": 2.6581,
+      "step": 171100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.455011207172591e-05,
+      "loss": 2.6755,
+      "step": 171200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.454691002241435e-05,
+      "loss": 2.6797,
+      "step": 171300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.454370797310279e-05,
+      "loss": 2.6632,
+      "step": 171400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.454050592379123e-05,
+      "loss": 2.6746,
+      "step": 171500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.453730387447967e-05,
+      "loss": 2.6487,
+      "step": 171600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4534101825168114e-05,
+      "loss": 2.6976,
+      "step": 171700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4530899775856547e-05,
+      "loss": 2.6726,
+      "step": 171800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.452769772654499e-05,
+      "loss": 2.6924,
+      "step": 171900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.452449567723343e-05,
+      "loss": 2.6741,
+      "step": 172000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.542654514312744,
+      "eval_runtime": 178.2761,
+      "eval_samples_per_second": 56.093,
+      "eval_steps_per_second": 3.506,
+      "step": 172000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.452129362792187e-05,
+      "loss": 2.6608,
+      "step": 172100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.451809157861031e-05,
+      "loss": 2.6683,
+      "step": 172200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.451488952929875e-05,
+      "loss": 2.6632,
+      "step": 172300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.45116874799872e-05,
+      "loss": 2.6721,
+      "step": 172400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.450848543067563e-05,
+      "loss": 2.6688,
+      "step": 172500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.450528338136408e-05,
+      "loss": 2.6955,
+      "step": 172600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.450208133205251e-05,
+      "loss": 2.6606,
+      "step": 172700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.449887928274096e-05,
+      "loss": 2.6826,
+      "step": 172800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.44956772334294e-05,
+      "loss": 2.681,
+      "step": 172900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.449247518411784e-05,
+      "loss": 2.6699,
+      "step": 173000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.5460827350616455,
+      "eval_runtime": 181.5889,
+      "eval_samples_per_second": 55.069,
+      "eval_steps_per_second": 3.442,
+      "step": 173000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4489273134806283e-05,
+      "loss": 2.6707,
+      "step": 173100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4486071085494716e-05,
+      "loss": 2.6739,
+      "step": 173200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.448286903618316e-05,
+      "loss": 2.671,
+      "step": 173300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4479666986871596e-05,
+      "loss": 2.678,
+      "step": 173400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.447646493756004e-05,
+      "loss": 2.6606,
+      "step": 173500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.447326288824848e-05,
+      "loss": 2.6661,
+      "step": 173600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.447006083893692e-05,
+      "loss": 2.6721,
+      "step": 173700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.446685878962536e-05,
+      "loss": 2.681,
+      "step": 173800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.44636567403138e-05,
+      "loss": 2.6674,
+      "step": 173900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.446045469100225e-05,
+      "loss": 2.6802,
+      "step": 174000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.543893814086914,
+      "eval_runtime": 179.174,
+      "eval_samples_per_second": 55.812,
+      "eval_steps_per_second": 3.488,
+      "step": 174000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.445725264169068e-05,
+      "loss": 2.6419,
+      "step": 174100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.445405059237913e-05,
+      "loss": 2.6505,
+      "step": 174200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.445084854306756e-05,
+      "loss": 2.6707,
+      "step": 174300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.444764649375601e-05,
+      "loss": 2.6756,
+      "step": 174400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 2.6702,
+      "step": 174500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4441242395132886e-05,
+      "loss": 2.6939,
+      "step": 174600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.443804034582133e-05,
+      "loss": 2.6484,
+      "step": 174700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4434838296509766e-05,
+      "loss": 2.6682,
+      "step": 174800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.443163624719821e-05,
+      "loss": 2.6815,
+      "step": 174900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4428434197886645e-05,
+      "loss": 2.6598,
+      "step": 175000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.5423295497894287,
+      "eval_runtime": 177.1753,
+      "eval_samples_per_second": 56.441,
+      "eval_steps_per_second": 3.528,
+      "step": 175000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.442523214857509e-05,
+      "loss": 2.6544,
+      "step": 175100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.442203009926353e-05,
+      "loss": 2.6505,
+      "step": 175200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.441882804995197e-05,
+      "loss": 2.6636,
+      "step": 175300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.441562600064042e-05,
+      "loss": 2.6802,
+      "step": 175400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.441242395132885e-05,
+      "loss": 2.6661,
+      "step": 175500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.44092219020173e-05,
+      "loss": 2.6685,
+      "step": 175600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.440601985270573e-05,
+      "loss": 2.6765,
+      "step": 175700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.440281780339418e-05,
+      "loss": 2.6613,
+      "step": 175800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.439961575408261e-05,
+      "loss": 2.6562,
+      "step": 175900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4396413704771056e-05,
+      "loss": 2.6664,
+      "step": 176000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.5409162044525146,
+      "eval_runtime": 177.764,
+      "eval_samples_per_second": 56.254,
+      "eval_steps_per_second": 3.516,
+      "step": 176000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4393211655459496e-05,
+      "loss": 2.6664,
+      "step": 176100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4390009606147936e-05,
+      "loss": 2.6751,
+      "step": 176200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.438680755683638e-05,
+      "loss": 2.6666,
+      "step": 176300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4383605507524815e-05,
+      "loss": 2.6851,
+      "step": 176400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.438040345821326e-05,
+      "loss": 2.6791,
+      "step": 176500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4377201408901695e-05,
+      "loss": 2.6624,
+      "step": 176600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.437399935959014e-05,
+      "loss": 2.6496,
+      "step": 176700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.437079731027858e-05,
+      "loss": 2.6762,
+      "step": 176800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.436759526096702e-05,
+      "loss": 2.6605,
+      "step": 176900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.436439321165547e-05,
+      "loss": 2.6654,
+      "step": 177000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.543023109436035,
+      "eval_runtime": 176.7668,
+      "eval_samples_per_second": 56.572,
+      "eval_steps_per_second": 3.536,
+      "step": 177000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.43611911623439e-05,
+      "loss": 2.6607,
+      "step": 177100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.435798911303235e-05,
+      "loss": 2.6708,
+      "step": 177200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.435478706372078e-05,
+      "loss": 2.6716,
+      "step": 177300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4351585014409226e-05,
+      "loss": 2.6454,
+      "step": 177400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.434838296509766e-05,
+      "loss": 2.6601,
+      "step": 177500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4345180915786106e-05,
+      "loss": 2.663,
+      "step": 177600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4341978866474545e-05,
+      "loss": 2.6717,
+      "step": 177700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4338776817162985e-05,
+      "loss": 2.6805,
+      "step": 177800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.433557476785143e-05,
+      "loss": 2.6775,
+      "step": 177900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4332372718539865e-05,
+      "loss": 2.6619,
+      "step": 178000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.539804458618164,
+      "eval_runtime": 176.7894,
+      "eval_samples_per_second": 56.564,
+      "eval_steps_per_second": 3.535,
+      "step": 178000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.432917066922831e-05,
+      "loss": 2.6875,
+      "step": 178100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4325968619916744e-05,
+      "loss": 2.663,
+      "step": 178200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.432276657060519e-05,
+      "loss": 2.6482,
+      "step": 178300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.431956452129363e-05,
+      "loss": 2.6495,
+      "step": 178400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.431636247198207e-05,
+      "loss": 2.6604,
+      "step": 178500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4313160422670517e-05,
+      "loss": 2.6784,
+      "step": 178600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.430995837335895e-05,
+      "loss": 2.6695,
+      "step": 178700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4306756324047396e-05,
+      "loss": 2.6449,
+      "step": 178800
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.430355427473583e-05,
+      "loss": 2.6526,
+      "step": 178900
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4300352225424275e-05,
+      "loss": 2.6478,
+      "step": 179000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.5400795936584473,
+      "eval_runtime": 179.3211,
+      "eval_samples_per_second": 55.766,
+      "eval_steps_per_second": 3.485,
+      "step": 179000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.429715017611271e-05,
+      "loss": 2.6662,
+      "step": 179100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4293948126801155e-05,
+      "loss": 2.6685,
+      "step": 179200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4290746077489595e-05,
+      "loss": 2.6561,
+      "step": 179300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4287544028178034e-05,
+      "loss": 2.6846,
+      "step": 179400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.428434197886648e-05,
+      "loss": 2.6364,
+      "step": 179500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4281139929554914e-05,
+      "loss": 2.6593,
+      "step": 179600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.427793788024336e-05,
+      "loss": 2.6733,
+      "step": 179700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.427473583093179e-05,
+      "loss": 2.6576,
+      "step": 179800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.427153378162024e-05,
+      "loss": 2.6718,
+      "step": 179900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.426833173230868e-05,
+      "loss": 2.6503,
+      "step": 180000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.541757345199585,
+      "eval_runtime": 178.2308,
+      "eval_samples_per_second": 56.107,
+      "eval_steps_per_second": 3.507,
+      "step": 180000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.426512968299712e-05,
+      "loss": 2.6436,
+      "step": 180100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4261927633685566e-05,
+      "loss": 2.6609,
+      "step": 180200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4258725584374e-05,
+      "loss": 2.6725,
+      "step": 180300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4255523535062445e-05,
+      "loss": 2.6301,
+      "step": 180400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.425232148575088e-05,
+      "loss": 2.6516,
+      "step": 180500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4249119436439325e-05,
+      "loss": 2.6533,
+      "step": 180600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4245917387127765e-05,
+      "loss": 2.6684,
+      "step": 180700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4242715337816204e-05,
+      "loss": 2.6714,
+      "step": 180800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4239513288504644e-05,
+      "loss": 2.6481,
+      "step": 180900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4236311239193084e-05,
+      "loss": 2.6519,
+      "step": 181000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5420303344726562,
+      "eval_runtime": 179.5147,
+      "eval_samples_per_second": 55.706,
+      "eval_steps_per_second": 3.482,
+      "step": 181000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.423310918988153e-05,
+      "loss": 2.6791,
+      "step": 181100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.422990714056996e-05,
+      "loss": 2.6628,
+      "step": 181200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.422670509125841e-05,
+      "loss": 2.6631,
+      "step": 181300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.422350304194684e-05,
+      "loss": 2.6562,
+      "step": 181400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.422030099263529e-05,
+      "loss": 2.6672,
+      "step": 181500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.421709894332373e-05,
+      "loss": 2.6597,
+      "step": 181600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.421389689401217e-05,
+      "loss": 2.6623,
+      "step": 181700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4210694844700615e-05,
+      "loss": 2.6592,
+      "step": 181800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.420749279538905e-05,
+      "loss": 2.6635,
+      "step": 181900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4204290746077495e-05,
+      "loss": 2.6597,
+      "step": 182000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.539924144744873,
+      "eval_runtime": 179.778,
+      "eval_samples_per_second": 55.624,
+      "eval_steps_per_second": 3.477,
+      "step": 182000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.420108869676593e-05,
+      "loss": 2.6532,
+      "step": 182100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4197886647454374e-05,
+      "loss": 2.6642,
+      "step": 182200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4194684598142814e-05,
+      "loss": 2.6746,
+      "step": 182300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4191482548831254e-05,
+      "loss": 2.6647,
+      "step": 182400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4188280499519693e-05,
+      "loss": 2.6517,
+      "step": 182500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.418507845020813e-05,
+      "loss": 2.6666,
+      "step": 182600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.418187640089658e-05,
+      "loss": 2.6635,
+      "step": 182700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.417867435158501e-05,
+      "loss": 2.7006,
+      "step": 182800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.417547230227346e-05,
+      "loss": 2.6571,
+      "step": 182900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.41722702529619e-05,
+      "loss": 2.6557,
+      "step": 183000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5385019779205322,
+      "eval_runtime": 177.971,
+      "eval_samples_per_second": 56.189,
+      "eval_steps_per_second": 3.512,
+      "step": 183000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.416906820365034e-05,
+      "loss": 2.6473,
+      "step": 183100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.416586615433878e-05,
+      "loss": 2.671,
+      "step": 183200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.416266410502722e-05,
+      "loss": 2.6507,
+      "step": 183300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4159462055715665e-05,
+      "loss": 2.6854,
+      "step": 183400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.41562600064041e-05,
+      "loss": 2.6632,
+      "step": 183500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4153057957092544e-05,
+      "loss": 2.6713,
+      "step": 183600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.414985590778098e-05,
+      "loss": 2.6622,
+      "step": 183700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4146653858469424e-05,
+      "loss": 2.6379,
+      "step": 183800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.414345180915786e-05,
+      "loss": 2.6656,
+      "step": 183900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.41402497598463e-05,
+      "loss": 2.6544,
+      "step": 184000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5391104221343994,
+      "eval_runtime": 182.9166,
+      "eval_samples_per_second": 54.67,
+      "eval_steps_per_second": 3.417,
+      "step": 184000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.413704771053474e-05,
+      "loss": 2.649,
+      "step": 184100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.413384566122318e-05,
+      "loss": 2.6583,
+      "step": 184200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.413064361191163e-05,
+      "loss": 2.6847,
+      "step": 184300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.412744156260006e-05,
+      "loss": 2.6579,
+      "step": 184400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.412423951328851e-05,
+      "loss": 2.6357,
+      "step": 184500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.412103746397695e-05,
+      "loss": 2.6437,
+      "step": 184600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.411783541466539e-05,
+      "loss": 2.6591,
+      "step": 184700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.411463336535383e-05,
+      "loss": 2.6783,
+      "step": 184800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.411143131604227e-05,
+      "loss": 2.6516,
+      "step": 184900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4108229266730714e-05,
+      "loss": 2.6584,
+      "step": 185000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5388665199279785,
+      "eval_runtime": 181.9145,
+      "eval_samples_per_second": 54.971,
+      "eval_steps_per_second": 3.436,
+      "step": 185000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.410502721741915e-05,
+      "loss": 2.6592,
+      "step": 185100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4101825168107593e-05,
+      "loss": 2.6676,
+      "step": 185200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.409862311879603e-05,
+      "loss": 2.6503,
+      "step": 185300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.409542106948447e-05,
+      "loss": 2.6789,
+      "step": 185400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.409221902017291e-05,
+      "loss": 2.6606,
+      "step": 185500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.408901697086135e-05,
+      "loss": 2.6591,
+      "step": 185600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.408581492154979e-05,
+      "loss": 2.6723,
+      "step": 185700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.408261287223823e-05,
+      "loss": 2.6434,
+      "step": 185800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.407941082292668e-05,
+      "loss": 2.6468,
+      "step": 185900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.407620877361511e-05,
+      "loss": 2.6492,
+      "step": 186000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.53727650642395,
+      "eval_runtime": 177.1235,
+      "eval_samples_per_second": 56.458,
+      "eval_steps_per_second": 3.529,
+      "step": 186000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.407300672430356e-05,
+      "loss": 2.6555,
+      "step": 186100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4069804674992e-05,
+      "loss": 2.6451,
+      "step": 186200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.406660262568044e-05,
+      "loss": 2.6524,
+      "step": 186300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.406340057636888e-05,
+      "loss": 2.6685,
+      "step": 186400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.406019852705732e-05,
+      "loss": 2.67,
+      "step": 186500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.405699647774576e-05,
+      "loss": 2.6588,
+      "step": 186600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4053794428434196e-05,
+      "loss": 2.6536,
+      "step": 186700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.405059237912264e-05,
+      "loss": 2.6571,
+      "step": 186800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.404739032981108e-05,
+      "loss": 2.6535,
+      "step": 186900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.404418828049952e-05,
+      "loss": 2.6514,
+      "step": 187000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5385489463806152,
+      "eval_runtime": 176.1879,
+      "eval_samples_per_second": 56.758,
+      "eval_steps_per_second": 3.547,
+      "step": 187000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.404098623118796e-05,
+      "loss": 2.6707,
+      "step": 187100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.40377841818764e-05,
+      "loss": 2.6565,
+      "step": 187200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.403458213256484e-05,
+      "loss": 2.65,
+      "step": 187300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.403138008325328e-05,
+      "loss": 2.6506,
+      "step": 187400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.402817803394173e-05,
+      "loss": 2.6517,
+      "step": 187500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.402497598463017e-05,
+      "loss": 2.6659,
+      "step": 187600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.402177393531861e-05,
+      "loss": 2.6645,
+      "step": 187700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.401857188600705e-05,
+      "loss": 2.6607,
+      "step": 187800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.401536983669549e-05,
+      "loss": 2.6688,
+      "step": 187900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4012167787383926e-05,
+      "loss": 2.6649,
+      "step": 188000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.537940740585327,
+      "eval_runtime": 179.0706,
+      "eval_samples_per_second": 55.844,
+      "eval_steps_per_second": 3.49,
+      "step": 188000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4008965738072366e-05,
+      "loss": 2.6627,
+      "step": 188100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4005763688760806e-05,
+      "loss": 2.6571,
+      "step": 188200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.400256163944925e-05,
+      "loss": 2.661,
+      "step": 188300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.399935959013769e-05,
+      "loss": 2.627,
+      "step": 188400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.399615754082613e-05,
+      "loss": 2.6708,
+      "step": 188500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.399295549151457e-05,
+      "loss": 2.6759,
+      "step": 188600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.398975344220301e-05,
+      "loss": 2.6571,
+      "step": 188700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.398655139289145e-05,
+      "loss": 2.673,
+      "step": 188800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.398334934357989e-05,
+      "loss": 2.667,
+      "step": 188900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.398014729426833e-05,
+      "loss": 2.6433,
+      "step": 189000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.539026975631714,
+      "eval_runtime": 176.2255,
+      "eval_samples_per_second": 56.745,
+      "eval_steps_per_second": 3.547,
+      "step": 189000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.397694524495678e-05,
+      "loss": 2.6545,
+      "step": 189100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.397374319564522e-05,
+      "loss": 2.6561,
+      "step": 189200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.397054114633366e-05,
+      "loss": 2.6411,
+      "step": 189300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3967339097022096e-05,
+      "loss": 2.6798,
+      "step": 189400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3964137047710536e-05,
+      "loss": 2.6658,
+      "step": 189500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3960934998398976e-05,
+      "loss": 2.6477,
+      "step": 189600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3957732949087416e-05,
+      "loss": 2.6476,
+      "step": 189700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3954530899775855e-05,
+      "loss": 2.6484,
+      "step": 189800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.39513288504643e-05,
+      "loss": 2.666,
+      "step": 189900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.394812680115274e-05,
+      "loss": 2.6422,
+      "step": 190000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.538339138031006,
+      "eval_runtime": 177.1468,
+      "eval_samples_per_second": 56.45,
+      "eval_steps_per_second": 3.528,
+      "step": 190000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.394492475184118e-05,
+      "loss": 2.6582,
+      "step": 190100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.394172270252962e-05,
+      "loss": 2.6722,
+      "step": 190200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.393852065321806e-05,
+      "loss": 2.6741,
+      "step": 190300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.39353186039065e-05,
+      "loss": 2.6614,
+      "step": 190400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.393211655459494e-05,
+      "loss": 2.6514,
+      "step": 190500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.392891450528339e-05,
+      "loss": 2.6624,
+      "step": 190600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3925712455971827e-05,
+      "loss": 2.6354,
+      "step": 190700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3922510406660266e-05,
+      "loss": 2.6575,
+      "step": 190800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3919308357348706e-05,
+      "loss": 2.6565,
+      "step": 190900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3916106308037146e-05,
+      "loss": 2.6553,
+      "step": 191000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5357167720794678,
+      "eval_runtime": 179.655,
+      "eval_samples_per_second": 55.662,
+      "eval_steps_per_second": 3.479,
+      "step": 191000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3912904258725585e-05,
+      "loss": 2.6707,
+      "step": 191100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3909702209414025e-05,
+      "loss": 2.6394,
+      "step": 191200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3906500160102465e-05,
+      "loss": 2.6648,
+      "step": 191300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3903298110790905e-05,
+      "loss": 2.6524,
+      "step": 191400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.390009606147935e-05,
+      "loss": 2.6517,
+      "step": 191500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.389689401216779e-05,
+      "loss": 2.6365,
+      "step": 191600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.389369196285623e-05,
+      "loss": 2.6565,
+      "step": 191700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.389048991354467e-05,
+      "loss": 2.6446,
+      "step": 191800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.388728786423311e-05,
+      "loss": 2.6494,
+      "step": 191900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.388408581492155e-05,
+      "loss": 2.6612,
+      "step": 192000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5360355377197266,
+      "eval_runtime": 177.2658,
+      "eval_samples_per_second": 56.412,
+      "eval_steps_per_second": 3.526,
+      "step": 192000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.388088376560999e-05,
+      "loss": 2.6504,
+      "step": 192100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3877681716298436e-05,
+      "loss": 2.634,
+      "step": 192200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3874479666986876e-05,
+      "loss": 2.6451,
+      "step": 192300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3871277617675316e-05,
+      "loss": 2.6446,
+      "step": 192400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3868075568363755e-05,
+      "loss": 2.6578,
+      "step": 192500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3864873519052195e-05,
+      "loss": 2.6589,
+      "step": 192600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3861671469740635e-05,
+      "loss": 2.659,
+      "step": 192700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3858469420429075e-05,
+      "loss": 2.6478,
+      "step": 192800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.385526737111752e-05,
+      "loss": 2.6644,
+      "step": 192900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3852065321805954e-05,
+      "loss": 2.6528,
+      "step": 193000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5370066165924072,
+      "eval_runtime": 176.8095,
+      "eval_samples_per_second": 56.558,
+      "eval_steps_per_second": 3.535,
+      "step": 193000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.38488632724944e-05,
+      "loss": 2.641,
+      "step": 193100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.384566122318284e-05,
+      "loss": 2.6679,
+      "step": 193200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.384245917387128e-05,
+      "loss": 2.6377,
+      "step": 193300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.383925712455972e-05,
+      "loss": 2.6468,
+      "step": 193400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.383605507524816e-05,
+      "loss": 2.6512,
+      "step": 193500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.38328530259366e-05,
+      "loss": 2.6409,
+      "step": 193600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.382965097662504e-05,
+      "loss": 2.6594,
+      "step": 193700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3826448927313486e-05,
+      "loss": 2.6408,
+      "step": 193800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3823246878001925e-05,
+      "loss": 2.6558,
+      "step": 193900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3820044828690365e-05,
+      "loss": 2.6412,
+      "step": 194000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.5341765880584717,
+      "eval_runtime": 176.7404,
+      "eval_samples_per_second": 56.58,
+      "eval_steps_per_second": 3.536,
+      "step": 194000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3816842779378805e-05,
+      "loss": 2.659,
+      "step": 194100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3813640730067245e-05,
+      "loss": 2.6463,
+      "step": 194200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3810438680755684e-05,
+      "loss": 2.6407,
+      "step": 194300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3807236631444124e-05,
+      "loss": 2.631,
+      "step": 194400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.380403458213257e-05,
+      "loss": 2.6573,
+      "step": 194500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3800832532821003e-05,
+      "loss": 2.6519,
+      "step": 194600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.379763048350945e-05,
+      "loss": 2.6476,
+      "step": 194700
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.379442843419789e-05,
+      "loss": 2.6345,
+      "step": 194800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.379122638488633e-05,
+      "loss": 2.6486,
+      "step": 194900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.378802433557477e-05,
+      "loss": 2.6419,
+      "step": 195000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.535737991333008,
+      "eval_runtime": 177.7521,
+      "eval_samples_per_second": 56.258,
+      "eval_steps_per_second": 3.516,
+      "step": 195000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.378482228626321e-05,
+      "loss": 2.6427,
+      "step": 195100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.3781620236951655e-05,
+      "loss": 2.651,
+      "step": 195200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.377841818764009e-05,
+      "loss": 2.6583,
+      "step": 195300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3775216138328535e-05,
+      "loss": 2.6303,
+      "step": 195400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3772014089016975e-05,
+      "loss": 2.6712,
+      "step": 195500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3768812039705414e-05,
+      "loss": 2.6521,
+      "step": 195600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3765609990393854e-05,
+      "loss": 2.6526,
+      "step": 195700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3762407941082294e-05,
+      "loss": 2.6319,
+      "step": 195800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3759205891770734e-05,
+      "loss": 2.6545,
+      "step": 195900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.375600384245917e-05,
+      "loss": 2.6323,
+      "step": 196000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.5361852645874023,
+      "eval_runtime": 178.5034,
+      "eval_samples_per_second": 56.021,
+      "eval_steps_per_second": 3.501,
+      "step": 196000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.375280179314762e-05,
+      "loss": 2.6563,
+      "step": 196100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.374959974383605e-05,
+      "loss": 2.6301,
+      "step": 196200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.37463976945245e-05,
+      "loss": 2.6575,
+      "step": 196300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.374319564521294e-05,
+      "loss": 2.6323,
+      "step": 196400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.373999359590138e-05,
+      "loss": 2.6314,
+      "step": 196500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.373679154658982e-05,
+      "loss": 2.6519,
+      "step": 196600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.373358949727826e-05,
+      "loss": 2.6487,
+      "step": 196700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3730387447966705e-05,
+      "loss": 2.6536,
+      "step": 196800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.372718539865514e-05,
+      "loss": 2.6511,
+      "step": 196900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3723983349343584e-05,
+      "loss": 2.6536,
+      "step": 197000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.533262252807617,
+      "eval_runtime": 177.806,
+      "eval_samples_per_second": 56.241,
+      "eval_steps_per_second": 3.515,
+      "step": 197000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3720781300032024e-05,
+      "loss": 2.672,
+      "step": 197100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3717579250720464e-05,
+      "loss": 2.655,
+      "step": 197200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3714377201408904e-05,
+      "loss": 2.6523,
+      "step": 197300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.371117515209734e-05,
+      "loss": 2.6679,
+      "step": 197400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.370797310278579e-05,
+      "loss": 2.6668,
+      "step": 197500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.370477105347422e-05,
+      "loss": 2.6433,
+      "step": 197600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.370156900416267e-05,
+      "loss": 2.66,
+      "step": 197700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.36983669548511e-05,
+      "loss": 2.6406,
+      "step": 197800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.369516490553955e-05,
+      "loss": 2.6566,
+      "step": 197900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.369196285622799e-05,
+      "loss": 2.6576,
+      "step": 198000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.53364634513855,
+      "eval_runtime": 175.0111,
+      "eval_samples_per_second": 57.139,
+      "eval_steps_per_second": 3.571,
+      "step": 198000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.368876080691643e-05,
+      "loss": 2.6356,
+      "step": 198100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3685558757604875e-05,
+      "loss": 2.663,
+      "step": 198200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.368235670829331e-05,
+      "loss": 2.6496,
+      "step": 198300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3679154658981754e-05,
+      "loss": 2.6501,
+      "step": 198400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.367595260967019e-05,
+      "loss": 2.6726,
+      "step": 198500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3672750560358634e-05,
+      "loss": 2.6322,
+      "step": 198600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3669548511047073e-05,
+      "loss": 2.6449,
+      "step": 198700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.366634646173551e-05,
+      "loss": 2.6556,
+      "step": 198800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.366314441242395e-05,
+      "loss": 2.6442,
+      "step": 198900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.365994236311239e-05,
+      "loss": 2.6787,
+      "step": 199000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.5323338508605957,
+      "eval_runtime": 177.6549,
+      "eval_samples_per_second": 56.289,
+      "eval_steps_per_second": 3.518,
+      "step": 199000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.365674031380084e-05,
+      "loss": 2.6481,
+      "step": 199100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.365353826448927e-05,
+      "loss": 2.6517,
+      "step": 199200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.365033621517772e-05,
+      "loss": 2.6406,
+      "step": 199300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.364713416586615e-05,
+      "loss": 2.6472,
+      "step": 199400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.36439321165546e-05,
+      "loss": 2.6541,
+      "step": 199500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.364073006724304e-05,
+      "loss": 2.6353,
+      "step": 199600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.363752801793148e-05,
+      "loss": 2.6538,
+      "step": 199700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3634325968619924e-05,
+      "loss": 2.6368,
+      "step": 199800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.363112391930836e-05,
+      "loss": 2.6507,
+      "step": 199900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3627921869996804e-05,
+      "loss": 2.6438,
+      "step": 200000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.533414125442505,
+      "eval_runtime": 176.5588,
+      "eval_samples_per_second": 56.638,
+      "eval_steps_per_second": 3.54,
+      "step": 200000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3624719820685237e-05,
+      "loss": 2.6773,
+      "step": 200100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.362151777137368e-05,
+      "loss": 2.6444,
+      "step": 200200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.361831572206212e-05,
+      "loss": 2.6491,
+      "step": 200300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.361511367275056e-05,
+      "loss": 2.6575,
+      "step": 200400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.361191162343901e-05,
+      "loss": 2.6546,
+      "step": 200500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.360870957412744e-05,
+      "loss": 2.6145,
+      "step": 200600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.360550752481589e-05,
+      "loss": 2.6463,
+      "step": 200700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.360230547550432e-05,
+      "loss": 2.6528,
+      "step": 200800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.359910342619277e-05,
+      "loss": 2.658,
+      "step": 200900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.35959013768812e-05,
+      "loss": 2.6667,
+      "step": 201000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.531358242034912,
+      "eval_runtime": 177.6252,
+      "eval_samples_per_second": 56.298,
+      "eval_steps_per_second": 3.519,
+      "step": 201000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.359269932756965e-05,
+      "loss": 2.6399,
+      "step": 201100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.358949727825809e-05,
+      "loss": 2.6425,
+      "step": 201200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.358629522894653e-05,
+      "loss": 2.6418,
+      "step": 201300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3583093179634973e-05,
+      "loss": 2.6688,
+      "step": 201400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3579891130323406e-05,
+      "loss": 2.6664,
+      "step": 201500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.357668908101185e-05,
+      "loss": 2.6546,
+      "step": 201600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3573487031700286e-05,
+      "loss": 2.6256,
+      "step": 201700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.357028498238873e-05,
+      "loss": 2.6314,
+      "step": 201800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.356708293307717e-05,
+      "loss": 2.6514,
+      "step": 201900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.356388088376561e-05,
+      "loss": 2.645,
+      "step": 202000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.5308117866516113,
+      "eval_runtime": 177.6308,
+      "eval_samples_per_second": 56.297,
+      "eval_steps_per_second": 3.519,
+      "step": 202000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.356067883445406e-05,
+      "loss": 2.6456,
+      "step": 202100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.355747678514249e-05,
+      "loss": 2.6435,
+      "step": 202200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.355427473583094e-05,
+      "loss": 2.6495,
+      "step": 202300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.355107268651937e-05,
+      "loss": 2.6596,
+      "step": 202400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.354787063720782e-05,
+      "loss": 2.6367,
+      "step": 202500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.354466858789625e-05,
+      "loss": 2.6302,
+      "step": 202600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.35414665385847e-05,
+      "loss": 2.6582,
+      "step": 202700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3538264489273137e-05,
+      "loss": 2.6292,
+      "step": 202800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3535062439961576e-05,
+      "loss": 2.651,
+      "step": 202900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.353186039065002e-05,
+      "loss": 2.6644,
+      "step": 203000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.53143572807312,
+      "eval_runtime": 175.7967,
+      "eval_samples_per_second": 56.884,
+      "eval_steps_per_second": 3.555,
+      "step": 203000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3528658341338456e-05,
+      "loss": 2.6447,
+      "step": 203100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.35254562920269e-05,
+      "loss": 2.6724,
+      "step": 203200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3522254242715335e-05,
+      "loss": 2.6471,
+      "step": 203300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.351905219340378e-05,
+      "loss": 2.6656,
+      "step": 203400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.351585014409222e-05,
+      "loss": 2.6418,
+      "step": 203500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.351264809478066e-05,
+      "loss": 2.6343,
+      "step": 203600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.350944604546911e-05,
+      "loss": 2.6404,
+      "step": 203700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.350624399615754e-05,
+      "loss": 2.666,
+      "step": 203800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.350304194684599e-05,
+      "loss": 2.6475,
+      "step": 203900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.349983989753442e-05,
+      "loss": 2.6607,
+      "step": 204000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.5313851833343506,
+      "eval_runtime": 176.4774,
+      "eval_samples_per_second": 56.664,
+      "eval_steps_per_second": 3.542,
+      "step": 204000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.349663784822287e-05,
+      "loss": 2.6304,
+      "step": 204100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.34934357989113e-05,
+      "loss": 2.655,
+      "step": 204200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3490233749599746e-05,
+      "loss": 2.6224,
+      "step": 204300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3487031700288186e-05,
+      "loss": 2.6145,
+      "step": 204400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3483829650976626e-05,
+      "loss": 2.6473,
+      "step": 204500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.348062760166507e-05,
+      "loss": 2.634,
+      "step": 204600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3477425552353505e-05,
+      "loss": 2.6438,
+      "step": 204700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.347422350304195e-05,
+      "loss": 2.6381,
+      "step": 204800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3471021453730385e-05,
+      "loss": 2.6346,
+      "step": 204900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.346781940441883e-05,
+      "loss": 2.633,
+      "step": 205000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.529742479324341,
+      "eval_runtime": 176.8338,
+      "eval_samples_per_second": 56.55,
+      "eval_steps_per_second": 3.534,
+      "step": 205000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.346461735510727e-05,
+      "loss": 2.6434,
+      "step": 205100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.346141530579571e-05,
+      "loss": 2.646,
+      "step": 205200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.345821325648416e-05,
+      "loss": 2.6348,
+      "step": 205300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.345501120717259e-05,
+      "loss": 2.632,
+      "step": 205400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.345180915786104e-05,
+      "loss": 2.6379,
+      "step": 205500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.344860710854947e-05,
+      "loss": 2.6433,
+      "step": 205600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3445405059237916e-05,
+      "loss": 2.632,
+      "step": 205700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3442203009926356e-05,
+      "loss": 2.6641,
+      "step": 205800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3439000960614796e-05,
+      "loss": 2.6363,
+      "step": 205900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3435798911303235e-05,
+      "loss": 2.6425,
+      "step": 206000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.528657913208008,
+      "eval_runtime": 175.9343,
+      "eval_samples_per_second": 56.839,
+      "eval_steps_per_second": 3.552,
+      "step": 206000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3432596861991675e-05,
+      "loss": 2.6231,
+      "step": 206100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.342939481268012e-05,
+      "loss": 2.6284,
+      "step": 206200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3426192763368555e-05,
+      "loss": 2.6581,
+      "step": 206300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3422990714057e-05,
+      "loss": 2.6255,
+      "step": 206400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3419788664745434e-05,
+      "loss": 2.652,
+      "step": 206500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.341658661543388e-05,
+      "loss": 2.632,
+      "step": 206600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.341338456612232e-05,
+      "loss": 2.6546,
+      "step": 206700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.341018251681076e-05,
+      "loss": 2.6312,
+      "step": 206800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.34069804674992e-05,
+      "loss": 2.6419,
+      "step": 206900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.340377841818764e-05,
+      "loss": 2.6417,
+      "step": 207000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.529862403869629,
+      "eval_runtime": 177.3132,
+      "eval_samples_per_second": 56.397,
+      "eval_steps_per_second": 3.525,
+      "step": 207000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3400576368876086e-05,
+      "loss": 2.645,
+      "step": 207100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.339737431956452e-05,
+      "loss": 2.6484,
+      "step": 207200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3394172270252965e-05,
+      "loss": 2.6235,
+      "step": 207300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3390970220941405e-05,
+      "loss": 2.633,
+      "step": 207400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3387768171629845e-05,
+      "loss": 2.6329,
+      "step": 207500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3384566122318285e-05,
+      "loss": 2.6663,
+      "step": 207600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3381364073006724e-05,
+      "loss": 2.6429,
+      "step": 207700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.337816202369517e-05,
+      "loss": 2.6539,
+      "step": 207800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3374959974383604e-05,
+      "loss": 2.6292,
+      "step": 207900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.337175792507205e-05,
+      "loss": 2.6195,
+      "step": 208000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.530228614807129,
+      "eval_runtime": 176.3625,
+      "eval_samples_per_second": 56.701,
+      "eval_steps_per_second": 3.544,
+      "step": 208000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.336855587576049e-05,
+      "loss": 2.6254,
+      "step": 208100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.336535382644893e-05,
+      "loss": 2.6354,
+      "step": 208200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.336215177713737e-05,
+      "loss": 2.6171,
+      "step": 208300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.335894972782581e-05,
+      "loss": 2.6298,
+      "step": 208400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.335574767851425e-05,
+      "loss": 2.6503,
+      "step": 208500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.335254562920269e-05,
+      "loss": 2.621,
+      "step": 208600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3349343579891135e-05,
+      "loss": 2.631,
+      "step": 208700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.334614153057957e-05,
+      "loss": 2.6321,
+      "step": 208800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3342939481268015e-05,
+      "loss": 2.6312,
+      "step": 208900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3339737431956455e-05,
+      "loss": 2.6512,
+      "step": 209000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.5286104679107666,
+      "eval_runtime": 195.3431,
+      "eval_samples_per_second": 51.192,
+      "eval_steps_per_second": 3.199,
+      "step": 209000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3336535382644894e-05,
+      "loss": 2.627,
+      "step": 209100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 2.6264,
+      "step": 209200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3330131284021774e-05,
+      "loss": 2.6263,
+      "step": 209300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.332692923471022e-05,
+      "loss": 2.6325,
+      "step": 209400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.332372718539865e-05,
+      "loss": 2.611,
+      "step": 209500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.33205251360871e-05,
+      "loss": 2.6166,
+      "step": 209600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.331732308677554e-05,
+      "loss": 2.6387,
+      "step": 209700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.331412103746398e-05,
+      "loss": 2.6505,
+      "step": 209800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.331091898815242e-05,
+      "loss": 2.6525,
+      "step": 209900
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.330771693884086e-05,
+      "loss": 2.6292,
+      "step": 210000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.5292112827301025,
+      "eval_runtime": 193.9899,
+      "eval_samples_per_second": 51.549,
+      "eval_steps_per_second": 3.222,
+      "step": 210000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.33045148895293e-05,
+      "loss": 2.646,
+      "step": 210100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.330131284021774e-05,
+      "loss": 2.617,
+      "step": 210200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3298110790906185e-05,
+      "loss": 2.6466,
+      "step": 210300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3294908741594624e-05,
+      "loss": 2.638,
+      "step": 210400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3291706692283064e-05,
+      "loss": 2.656,
+      "step": 210500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3288504642971504e-05,
+      "loss": 2.6414,
+      "step": 210600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3285302593659944e-05,
+      "loss": 2.629,
+      "step": 210700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3282100544348383e-05,
+      "loss": 2.6291,
+      "step": 210800
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.327889849503682e-05,
+      "loss": 2.6474,
+      "step": 210900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.327569644572527e-05,
+      "loss": 2.6521,
+      "step": 211000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.529031991958618,
+      "eval_runtime": 177.8606,
+      "eval_samples_per_second": 56.224,
+      "eval_steps_per_second": 3.514,
+      "step": 211000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.32724943964137e-05,
+      "loss": 2.6118,
+      "step": 211100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.326929234710215e-05,
+      "loss": 2.6543,
+      "step": 211200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.326609029779059e-05,
+      "loss": 2.6448,
+      "step": 211300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.326288824847903e-05,
+      "loss": 2.6423,
+      "step": 211400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.325968619916747e-05,
+      "loss": 2.6156,
+      "step": 211500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.325648414985591e-05,
+      "loss": 2.6427,
+      "step": 211600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.325328210054435e-05,
+      "loss": 2.6329,
+      "step": 211700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.325008005123279e-05,
+      "loss": 2.6331,
+      "step": 211800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3246878001921234e-05,
+      "loss": 2.6276,
+      "step": 211900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3243675952609674e-05,
+      "loss": 2.6377,
+      "step": 212000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5291922092437744,
+      "eval_runtime": 176.4283,
+      "eval_samples_per_second": 56.68,
+      "eval_steps_per_second": 3.543,
+      "step": 212000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3240473903298114e-05,
+      "loss": 2.6265,
+      "step": 212100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.323727185398655e-05,
+      "loss": 2.6319,
+      "step": 212200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.323406980467499e-05,
+      "loss": 2.6235,
+      "step": 212300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.323086775536343e-05,
+      "loss": 2.6233,
+      "step": 212400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.322766570605187e-05,
+      "loss": 2.6506,
+      "step": 212500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.322446365674032e-05,
+      "loss": 2.6294,
+      "step": 212600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.322126160742876e-05,
+      "loss": 2.6288,
+      "step": 212700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.32180595581172e-05,
+      "loss": 2.6241,
+      "step": 212800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.321485750880564e-05,
+      "loss": 2.6348,
+      "step": 212900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.321165545949408e-05,
+      "loss": 2.6268,
+      "step": 213000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5289108753204346,
+      "eval_runtime": 173.9573,
+      "eval_samples_per_second": 57.485,
+      "eval_steps_per_second": 3.593,
+      "step": 213000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.320845341018252e-05,
+      "loss": 2.5975,
+      "step": 213100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.320525136087096e-05,
+      "loss": 2.6618,
+      "step": 213200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.32020493115594e-05,
+      "loss": 2.6427,
+      "step": 213300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.319884726224784e-05,
+      "loss": 2.6347,
+      "step": 213400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3195645212936283e-05,
+      "loss": 2.6308,
+      "step": 213500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.319244316362472e-05,
+      "loss": 2.6144,
+      "step": 213600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.318924111431316e-05,
+      "loss": 2.6242,
+      "step": 213700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.31860390650016e-05,
+      "loss": 2.6297,
+      "step": 213800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.318283701569004e-05,
+      "loss": 2.6198,
+      "step": 213900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.317963496637848e-05,
+      "loss": 2.6436,
+      "step": 214000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.528165340423584,
+      "eval_runtime": 176.0011,
+      "eval_samples_per_second": 56.818,
+      "eval_steps_per_second": 3.551,
+      "step": 214000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.317643291706692e-05,
+      "loss": 2.6456,
+      "step": 214100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.317323086775537e-05,
+      "loss": 2.6205,
+      "step": 214200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.317002881844381e-05,
+      "loss": 2.6102,
+      "step": 214300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.316682676913225e-05,
+      "loss": 2.6279,
+      "step": 214400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.316362471982069e-05,
+      "loss": 2.6337,
+      "step": 214500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.316042267050913e-05,
+      "loss": 2.6296,
+      "step": 214600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.315722062119757e-05,
+      "loss": 2.6301,
+      "step": 214700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.315401857188601e-05,
+      "loss": 2.6218,
+      "step": 214800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3150816522574447e-05,
+      "loss": 2.6434,
+      "step": 214900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.314761447326289e-05,
+      "loss": 2.61,
+      "step": 215000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5294289588928223,
+      "eval_runtime": 177.7074,
+      "eval_samples_per_second": 56.272,
+      "eval_steps_per_second": 3.517,
+      "step": 215000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.314441242395133e-05,
+      "loss": 2.6298,
+      "step": 215100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.314121037463977e-05,
+      "loss": 2.6189,
+      "step": 215200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.313800832532821e-05,
+      "loss": 2.596,
+      "step": 215300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.313480627601665e-05,
+      "loss": 2.6313,
+      "step": 215400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.313160422670509e-05,
+      "loss": 2.6104,
+      "step": 215500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.312840217739353e-05,
+      "loss": 2.6391,
+      "step": 215600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.312520012808198e-05,
+      "loss": 2.6356,
+      "step": 215700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.312199807877042e-05,
+      "loss": 2.6151,
+      "step": 215800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.311879602945886e-05,
+      "loss": 2.6279,
+      "step": 215900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.31155939801473e-05,
+      "loss": 2.6185,
+      "step": 216000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5280823707580566,
+      "eval_runtime": 176.0196,
+      "eval_samples_per_second": 56.812,
+      "eval_steps_per_second": 3.551,
+      "step": 216000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.311239193083574e-05,
+      "loss": 2.6091,
+      "step": 216100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.310918988152418e-05,
+      "loss": 2.6187,
+      "step": 216200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3105987832212616e-05,
+      "loss": 2.6362,
+      "step": 216300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3102785782901056e-05,
+      "loss": 2.617,
+      "step": 216400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3099583733589496e-05,
+      "loss": 2.6403,
+      "step": 216500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.309638168427794e-05,
+      "loss": 2.6077,
+      "step": 216600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.309317963496638e-05,
+      "loss": 2.6381,
+      "step": 216700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.308997758565482e-05,
+      "loss": 2.6548,
+      "step": 216800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.308677553634326e-05,
+      "loss": 2.6257,
+      "step": 216900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.30835734870317e-05,
+      "loss": 2.6427,
+      "step": 217000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5252022743225098,
+      "eval_runtime": 173.6916,
+      "eval_samples_per_second": 57.573,
+      "eval_steps_per_second": 3.598,
+      "step": 217000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.308037143772014e-05,
+      "loss": 2.6324,
+      "step": 217100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.307716938840858e-05,
+      "loss": 2.62,
+      "step": 217200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.307396733909703e-05,
+      "loss": 2.6349,
+      "step": 217300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.307076528978547e-05,
+      "loss": 2.6162,
+      "step": 217400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.306756324047391e-05,
+      "loss": 2.6245,
+      "step": 217500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.306436119116235e-05,
+      "loss": 2.6265,
+      "step": 217600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3061159141850786e-05,
+      "loss": 2.6246,
+      "step": 217700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3057957092539226e-05,
+      "loss": 2.6485,
+      "step": 217800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3054755043227666e-05,
+      "loss": 2.6113,
+      "step": 217900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.305155299391611e-05,
+      "loss": 2.6212,
+      "step": 218000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.527156352996826,
+      "eval_runtime": 176.014,
+      "eval_samples_per_second": 56.814,
+      "eval_steps_per_second": 3.551,
+      "step": 218000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3048350944604545e-05,
+      "loss": 2.6248,
+      "step": 218100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.304514889529299e-05,
+      "loss": 2.6509,
+      "step": 218200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.304194684598143e-05,
+      "loss": 2.6011,
+      "step": 218300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.303874479666987e-05,
+      "loss": 2.6169,
+      "step": 218400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.303554274735831e-05,
+      "loss": 2.6246,
+      "step": 218500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.303234069804675e-05,
+      "loss": 2.6467,
+      "step": 218600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.302913864873519e-05,
+      "loss": 2.6275,
+      "step": 218700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.302593659942363e-05,
+      "loss": 2.6104,
+      "step": 218800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.302273455011208e-05,
+      "loss": 2.613,
+      "step": 218900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3019532500800517e-05,
+      "loss": 2.64,
+      "step": 219000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5278069972991943,
+      "eval_runtime": 176.722,
+      "eval_samples_per_second": 56.586,
+      "eval_steps_per_second": 3.537,
+      "step": 219000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3016330451488956e-05,
+      "loss": 2.6189,
+      "step": 219100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3013128402177396e-05,
+      "loss": 2.6284,
+      "step": 219200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3009926352865836e-05,
+      "loss": 2.6216,
+      "step": 219300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3006724303554275e-05,
+      "loss": 2.6361,
+      "step": 219400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3003522254242715e-05,
+      "loss": 2.6333,
+      "step": 219500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.300032020493116e-05,
+      "loss": 2.6239,
+      "step": 219600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2997118155619595e-05,
+      "loss": 2.6211,
+      "step": 219700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.299391610630804e-05,
+      "loss": 2.635,
+      "step": 219800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.299071405699648e-05,
+      "loss": 2.6111,
+      "step": 219900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.298751200768492e-05,
+      "loss": 2.6169,
+      "step": 220000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.527263641357422,
+      "eval_runtime": 173.9107,
+      "eval_samples_per_second": 57.501,
+      "eval_steps_per_second": 3.594,
+      "step": 220000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.298430995837336e-05,
+      "loss": 2.595,
+      "step": 220100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.29811079090618e-05,
+      "loss": 2.6117,
+      "step": 220200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.297790585975025e-05,
+      "loss": 2.635,
+      "step": 220300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.297470381043868e-05,
+      "loss": 2.6172,
+      "step": 220400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2971501761127126e-05,
+      "loss": 2.6418,
+      "step": 220500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2968299711815566e-05,
+      "loss": 2.6069,
+      "step": 220600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2965097662504006e-05,
+      "loss": 2.6341,
+      "step": 220700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2961895613192445e-05,
+      "loss": 2.6143,
+      "step": 220800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2958693563880885e-05,
+      "loss": 2.6295,
+      "step": 220900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2955491514569325e-05,
+      "loss": 2.6249,
+      "step": 221000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.526632308959961,
+      "eval_runtime": 178.088,
+      "eval_samples_per_second": 56.152,
+      "eval_steps_per_second": 3.51,
+      "step": 221000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2952289465257765e-05,
+      "loss": 2.614,
+      "step": 221100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.294908741594621e-05,
+      "loss": 2.6068,
+      "step": 221200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2945885366634644e-05,
+      "loss": 2.6442,
+      "step": 221300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.294268331732309e-05,
+      "loss": 2.6008,
+      "step": 221400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.293948126801153e-05,
+      "loss": 2.6161,
+      "step": 221500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.293627921869997e-05,
+      "loss": 2.6089,
+      "step": 221600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.293307716938841e-05,
+      "loss": 2.6385,
+      "step": 221700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.292987512007685e-05,
+      "loss": 2.6166,
+      "step": 221800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2926673070765296e-05,
+      "loss": 2.6013,
+      "step": 221900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.292347102145373e-05,
+      "loss": 2.6103,
+      "step": 222000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5263125896453857,
+      "eval_runtime": 178.1295,
+      "eval_samples_per_second": 56.139,
+      "eval_steps_per_second": 3.509,
+      "step": 222000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2920268972142176e-05,
+      "loss": 2.62,
+      "step": 222100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2917066922830615e-05,
+      "loss": 2.6053,
+      "step": 222200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2913864873519055e-05,
+      "loss": 2.6139,
+      "step": 222300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2910662824207495e-05,
+      "loss": 2.617,
+      "step": 222400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2907460774895934e-05,
+      "loss": 2.6101,
+      "step": 222500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.290425872558438e-05,
+      "loss": 2.6292,
+      "step": 222600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2901056676272814e-05,
+      "loss": 2.608,
+      "step": 222700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.289785462696126e-05,
+      "loss": 2.6042,
+      "step": 222800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2894652577649693e-05,
+      "loss": 2.5969,
+      "step": 222900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.289145052833814e-05,
+      "loss": 2.6138,
+      "step": 223000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.525189161300659,
+      "eval_runtime": 174.7154,
+      "eval_samples_per_second": 57.236,
+      "eval_steps_per_second": 3.577,
+      "step": 223000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.288824847902658e-05,
+      "loss": 2.6086,
+      "step": 223100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.288504642971502e-05,
+      "loss": 2.6301,
+      "step": 223200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.288184438040346e-05,
+      "loss": 2.6091,
+      "step": 223300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.28786423310919e-05,
+      "loss": 2.6136,
+      "step": 223400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2875440281780345e-05,
+      "loss": 2.5998,
+      "step": 223500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.287223823246878e-05,
+      "loss": 2.6226,
+      "step": 223600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2869036183157225e-05,
+      "loss": 2.6109,
+      "step": 223700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2865834133845665e-05,
+      "loss": 2.633,
+      "step": 223800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2862632084534104e-05,
+      "loss": 2.6114,
+      "step": 223900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2859430035222544e-05,
+      "loss": 2.6099,
+      "step": 224000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5254721641540527,
+      "eval_runtime": 176.1995,
+      "eval_samples_per_second": 56.754,
+      "eval_steps_per_second": 3.547,
+      "step": 224000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2856227985910984e-05,
+      "loss": 2.6315,
+      "step": 224100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.285302593659943e-05,
+      "loss": 2.6034,
+      "step": 224200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.284982388728786e-05,
+      "loss": 2.5927,
+      "step": 224300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.284662183797631e-05,
+      "loss": 2.6138,
+      "step": 224400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.284341978866474e-05,
+      "loss": 2.6295,
+      "step": 224500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.284021773935319e-05,
+      "loss": 2.6111,
+      "step": 224600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.283701569004163e-05,
+      "loss": 2.6108,
+      "step": 224700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.283381364073007e-05,
+      "loss": 2.6103,
+      "step": 224800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2830611591418515e-05,
+      "loss": 2.6197,
+      "step": 224900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.282740954210695e-05,
+      "loss": 2.6191,
+      "step": 225000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.5249056816101074,
+      "eval_runtime": 176.5788,
+      "eval_samples_per_second": 56.632,
+      "eval_steps_per_second": 3.539,
+      "step": 225000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2824207492795395e-05,
+      "loss": 2.6135,
+      "step": 225100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.282100544348383e-05,
+      "loss": 2.6273,
+      "step": 225200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2817803394172274e-05,
+      "loss": 2.6085,
+      "step": 225300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2814601344860714e-05,
+      "loss": 2.6303,
+      "step": 225400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2811399295549154e-05,
+      "loss": 2.609,
+      "step": 225500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2808197246237594e-05,
+      "loss": 2.6163,
+      "step": 225600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.280499519692603e-05,
+      "loss": 2.625,
+      "step": 225700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.280179314761448e-05,
+      "loss": 2.6277,
+      "step": 225800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.279859109830291e-05,
+      "loss": 2.628,
+      "step": 225900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.279538904899136e-05,
+      "loss": 2.6122,
+      "step": 226000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.527702808380127,
+      "eval_runtime": 179.746,
+      "eval_samples_per_second": 55.634,
+      "eval_steps_per_second": 3.477,
+      "step": 226000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.279218699967979e-05,
+      "loss": 2.6287,
+      "step": 226100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.278898495036824e-05,
+      "loss": 2.6267,
+      "step": 226200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.278578290105668e-05,
+      "loss": 2.6106,
+      "step": 226300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.278258085174512e-05,
+      "loss": 2.6114,
+      "step": 226400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.2779378802433565e-05,
+      "loss": 2.6216,
+      "step": 226500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2776176753122e-05,
+      "loss": 2.6045,
+      "step": 226600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2772974703810444e-05,
+      "loss": 2.6172,
+      "step": 226700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.276977265449888e-05,
+      "loss": 2.6315,
+      "step": 226800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2766570605187324e-05,
+      "loss": 2.6167,
+      "step": 226900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2763368555875763e-05,
+      "loss": 2.6426,
+      "step": 227000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.525062084197998,
+      "eval_runtime": 177.7314,
+      "eval_samples_per_second": 56.265,
+      "eval_steps_per_second": 3.517,
+      "step": 227000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.27601665065642e-05,
+      "loss": 2.6213,
+      "step": 227100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.275696445725264e-05,
+      "loss": 2.6122,
+      "step": 227200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.275376240794108e-05,
+      "loss": 2.6104,
+      "step": 227300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.275056035862953e-05,
+      "loss": 2.6016,
+      "step": 227400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.274735830931796e-05,
+      "loss": 2.5993,
+      "step": 227500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.274415626000641e-05,
+      "loss": 2.6125,
+      "step": 227600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.274095421069484e-05,
+      "loss": 2.5991,
+      "step": 227700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.273775216138329e-05,
+      "loss": 2.619,
+      "step": 227800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.273455011207173e-05,
+      "loss": 2.6267,
+      "step": 227900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.273134806276017e-05,
+      "loss": 2.613,
+      "step": 228000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.52404522895813,
+      "eval_runtime": 176.0414,
+      "eval_samples_per_second": 56.805,
+      "eval_steps_per_second": 3.55,
+      "step": 228000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2728146013448614e-05,
+      "loss": 2.626,
+      "step": 228100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.272494396413705e-05,
+      "loss": 2.6154,
+      "step": 228200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2721741914825494e-05,
+      "loss": 2.6089,
+      "step": 228300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2718539865513927e-05,
+      "loss": 2.6217,
+      "step": 228400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.271533781620237e-05,
+      "loss": 2.6165,
+      "step": 228500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.271213576689081e-05,
+      "loss": 2.6223,
+      "step": 228600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.270893371757925e-05,
+      "loss": 2.6154,
+      "step": 228700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.270573166826769e-05,
+      "loss": 2.6102,
+      "step": 228800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.270252961895613e-05,
+      "loss": 2.631,
+      "step": 228900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.269932756964458e-05,
+      "loss": 2.6001,
+      "step": 229000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.5241382122039795,
+      "eval_runtime": 176.3416,
+      "eval_samples_per_second": 56.708,
+      "eval_steps_per_second": 3.544,
+      "step": 229000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.269612552033301e-05,
+      "loss": 2.5996,
+      "step": 229100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.269292347102146e-05,
+      "loss": 2.6287,
+      "step": 229200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.268972142170989e-05,
+      "loss": 2.6207,
+      "step": 229300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.268651937239834e-05,
+      "loss": 2.6177,
+      "step": 229400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.268331732308678e-05,
+      "loss": 2.6006,
+      "step": 229500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.268011527377522e-05,
+      "loss": 2.6149,
+      "step": 229600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2676913224463663e-05,
+      "loss": 2.6005,
+      "step": 229700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2673711175152096e-05,
+      "loss": 2.6295,
+      "step": 229800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.267050912584054e-05,
+      "loss": 2.6113,
+      "step": 229900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2667307076528976e-05,
+      "loss": 2.6262,
+      "step": 230000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.5240397453308105,
+      "eval_runtime": 177.6269,
+      "eval_samples_per_second": 56.298,
+      "eval_steps_per_second": 3.519,
+      "step": 230000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.266410502721742e-05,
+      "loss": 2.6225,
+      "step": 230100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.266090297790586e-05,
+      "loss": 2.6236,
+      "step": 230200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.26577009285943e-05,
+      "loss": 2.6117,
+      "step": 230300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.265449887928274e-05,
+      "loss": 2.6258,
+      "step": 230400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.265129682997118e-05,
+      "loss": 2.6132,
+      "step": 230500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.264809478065963e-05,
+      "loss": 2.6055,
+      "step": 230600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.264489273134806e-05,
+      "loss": 2.6217,
+      "step": 230700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.264169068203651e-05,
+      "loss": 2.6134,
+      "step": 230800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.263848863272494e-05,
+      "loss": 2.6186,
+      "step": 230900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.263528658341339e-05,
+      "loss": 2.6124,
+      "step": 231000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.522747755050659,
+      "eval_runtime": 177.0914,
+      "eval_samples_per_second": 56.468,
+      "eval_steps_per_second": 3.529,
+      "step": 231000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2632084534101827e-05,
+      "loss": 2.6188,
+      "step": 231100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2628882484790266e-05,
+      "loss": 2.6348,
+      "step": 231200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.262568043547871e-05,
+      "loss": 2.628,
+      "step": 231300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2622478386167146e-05,
+      "loss": 2.6006,
+      "step": 231400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.261927633685559e-05,
+      "loss": 2.6075,
+      "step": 231500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2616074287544025e-05,
+      "loss": 2.6141,
+      "step": 231600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.261287223823247e-05,
+      "loss": 2.6051,
+      "step": 231700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.260967018892091e-05,
+      "loss": 2.6123,
+      "step": 231800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.260646813960935e-05,
+      "loss": 2.6122,
+      "step": 231900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.260326609029779e-05,
+      "loss": 2.6129,
+      "step": 232000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.5222604274749756,
+      "eval_runtime": 175.6813,
+      "eval_samples_per_second": 56.921,
+      "eval_steps_per_second": 3.558,
+      "step": 232000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.260006404098623e-05,
+      "loss": 2.6264,
+      "step": 232100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.259686199167468e-05,
+      "loss": 2.6076,
+      "step": 232200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.259365994236311e-05,
+      "loss": 2.6195,
+      "step": 232300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.259045789305156e-05,
+      "loss": 2.6187,
+      "step": 232400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2587255843739996e-05,
+      "loss": 2.6214,
+      "step": 232500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2584053794428436e-05,
+      "loss": 2.6129,
+      "step": 232600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2580851745116876e-05,
+      "loss": 2.6233,
+      "step": 232700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2577649695805316e-05,
+      "loss": 2.61,
+      "step": 232800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.257444764649376e-05,
+      "loss": 2.6174,
+      "step": 232900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2571245597182195e-05,
+      "loss": 2.6273,
+      "step": 233000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.52047061920166,
+      "eval_runtime": 175.8957,
+      "eval_samples_per_second": 56.852,
+      "eval_steps_per_second": 3.553,
+      "step": 233000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.256804354787064e-05,
+      "loss": 2.6267,
+      "step": 233100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2564841498559075e-05,
+      "loss": 2.6121,
+      "step": 233200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.256163944924752e-05,
+      "loss": 2.6026,
+      "step": 233300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.255843739993596e-05,
+      "loss": 2.6381,
+      "step": 233400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.25552353506244e-05,
+      "loss": 2.6196,
+      "step": 233500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.255203330131284e-05,
+      "loss": 2.6094,
+      "step": 233600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.254883125200128e-05,
+      "loss": 2.6105,
+      "step": 233700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.254562920268973e-05,
+      "loss": 2.6103,
+      "step": 233800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.254242715337816e-05,
+      "loss": 2.6288,
+      "step": 233900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2539225104066606e-05,
+      "loss": 2.6153,
+      "step": 234000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.5217788219451904,
+      "eval_runtime": 178.8277,
+      "eval_samples_per_second": 55.92,
+      "eval_steps_per_second": 3.495,
+      "step": 234000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2536023054755046e-05,
+      "loss": 2.6037,
+      "step": 234100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2532821005443486e-05,
+      "loss": 2.6315,
+      "step": 234200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2529618956131925e-05,
+      "loss": 2.6008,
+      "step": 234300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2526416906820365e-05,
+      "loss": 2.6132,
+      "step": 234400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.252321485750881e-05,
+      "loss": 2.5825,
+      "step": 234500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2520012808197245e-05,
+      "loss": 2.6072,
+      "step": 234600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.251681075888569e-05,
+      "loss": 2.6044,
+      "step": 234700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.251360870957413e-05,
+      "loss": 2.6369,
+      "step": 234800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.251040666026257e-05,
+      "loss": 2.6158,
+      "step": 234900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.250720461095101e-05,
+      "loss": 2.6374,
+      "step": 235000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.521732807159424,
+      "eval_runtime": 176.1112,
+      "eval_samples_per_second": 56.782,
+      "eval_steps_per_second": 3.549,
+      "step": 235000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.250400256163945e-05,
+      "loss": 2.6234,
+      "step": 235100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.250080051232789e-05,
+      "loss": 2.6115,
+      "step": 235200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.249759846301633e-05,
+      "loss": 2.6079,
+      "step": 235300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2494396413704776e-05,
+      "loss": 2.637,
+      "step": 235400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2491194364393216e-05,
+      "loss": 2.6268,
+      "step": 235500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2487992315081655e-05,
+      "loss": 2.6055,
+      "step": 235600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2484790265770095e-05,
+      "loss": 2.5941,
+      "step": 235700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2481588216458535e-05,
+      "loss": 2.6161,
+      "step": 235800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2478386167146975e-05,
+      "loss": 2.5807,
+      "step": 235900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2475184117835414e-05,
+      "loss": 2.6111,
+      "step": 236000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.522373676300049,
+      "eval_runtime": 177.1075,
+      "eval_samples_per_second": 56.463,
+      "eval_steps_per_second": 3.529,
+      "step": 236000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.247198206852386e-05,
+      "loss": 2.6346,
+      "step": 236100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2468780019212294e-05,
+      "loss": 2.6172,
+      "step": 236200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.246557796990074e-05,
+      "loss": 2.6092,
+      "step": 236300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.246237592058918e-05,
+      "loss": 2.619,
+      "step": 236400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.245917387127762e-05,
+      "loss": 2.6073,
+      "step": 236500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.245597182196606e-05,
+      "loss": 2.626,
+      "step": 236600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.24527697726545e-05,
+      "loss": 2.6131,
+      "step": 236700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.244956772334294e-05,
+      "loss": 2.5899,
+      "step": 236800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.244636567403138e-05,
+      "loss": 2.5973,
+      "step": 236900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2443163624719825e-05,
+      "loss": 2.6271,
+      "step": 237000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.5218992233276367,
+      "eval_runtime": 177.3058,
+      "eval_samples_per_second": 56.4,
+      "eval_steps_per_second": 3.525,
+      "step": 237000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2439961575408265e-05,
+      "loss": 2.6092,
+      "step": 237100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2436759526096705e-05,
+      "loss": 2.601,
+      "step": 237200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2433557476785145e-05,
+      "loss": 2.618,
+      "step": 237300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2430355427473584e-05,
+      "loss": 2.6241,
+      "step": 237400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2427153378162024e-05,
+      "loss": 2.6135,
+      "step": 237500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2423951328850464e-05,
+      "loss": 2.6208,
+      "step": 237600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.242074927953891e-05,
+      "loss": 2.6056,
+      "step": 237700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.241754723022735e-05,
+      "loss": 2.6001,
+      "step": 237800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.241434518091579e-05,
+      "loss": 2.6226,
+      "step": 237900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.241114313160423e-05,
+      "loss": 2.6156,
+      "step": 238000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.519456624984741,
+      "eval_runtime": 177.7406,
+      "eval_samples_per_second": 56.262,
+      "eval_steps_per_second": 3.516,
+      "step": 238000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.240794108229267e-05,
+      "loss": 2.6083,
+      "step": 238100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.240473903298111e-05,
+      "loss": 2.6022,
+      "step": 238200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.240153698366955e-05,
+      "loss": 2.6034,
+      "step": 238300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.239833493435799e-05,
+      "loss": 2.6088,
+      "step": 238400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.239513288504643e-05,
+      "loss": 2.6109,
+      "step": 238500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2391930835734875e-05,
+      "loss": 2.5811,
+      "step": 238600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2388728786423314e-05,
+      "loss": 2.61,
+      "step": 238700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2385526737111754e-05,
+      "loss": 2.6104,
+      "step": 238800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2382324687800194e-05,
+      "loss": 2.6341,
+      "step": 238900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2379122638488634e-05,
+      "loss": 2.6085,
+      "step": 239000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.521357536315918,
+      "eval_runtime": 180.2145,
+      "eval_samples_per_second": 55.489,
+      "eval_steps_per_second": 3.468,
+      "step": 239000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2375920589177073e-05,
+      "loss": 2.6155,
+      "step": 239100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.237271853986551e-05,
+      "loss": 2.62,
+      "step": 239200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.236951649055396e-05,
+      "loss": 2.6162,
+      "step": 239300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.23663144412424e-05,
+      "loss": 2.6319,
+      "step": 239400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.236311239193084e-05,
+      "loss": 2.5933,
+      "step": 239500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.235991034261928e-05,
+      "loss": 2.6053,
+      "step": 239600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.235670829330772e-05,
+      "loss": 2.6177,
+      "step": 239700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.235350624399616e-05,
+      "loss": 2.6148,
+      "step": 239800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.23503041946846e-05,
+      "loss": 2.6012,
+      "step": 239900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.234710214537304e-05,
+      "loss": 2.6172,
+      "step": 240000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.522383451461792,
+      "eval_runtime": 176.722,
+      "eval_samples_per_second": 56.586,
+      "eval_steps_per_second": 3.537,
+      "step": 240000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2343900096061484e-05,
+      "loss": 2.6025,
+      "step": 240100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2340698046749924e-05,
+      "loss": 2.6058,
+      "step": 240200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2337495997438364e-05,
+      "loss": 2.5907,
+      "step": 240300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2334293948126804e-05,
+      "loss": 2.6172,
+      "step": 240400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.233109189881524e-05,
+      "loss": 2.6035,
+      "step": 240500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.232788984950368e-05,
+      "loss": 2.6052,
+      "step": 240600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.232468780019212e-05,
+      "loss": 2.5891,
+      "step": 240700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.232148575088056e-05,
+      "loss": 2.6134,
+      "step": 240800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.231828370156901e-05,
+      "loss": 2.6096,
+      "step": 240900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.231508165225745e-05,
+      "loss": 2.6041,
+      "step": 241000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.519827127456665,
+      "eval_runtime": 173.9651,
+      "eval_samples_per_second": 57.483,
+      "eval_steps_per_second": 3.593,
+      "step": 241000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.231187960294589e-05,
+      "loss": 2.6267,
+      "step": 241100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.230867755363433e-05,
+      "loss": 2.6165,
+      "step": 241200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.230547550432277e-05,
+      "loss": 2.5852,
+      "step": 241300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.230227345501121e-05,
+      "loss": 2.5852,
+      "step": 241400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.229907140569965e-05,
+      "loss": 2.6022,
+      "step": 241500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.229586935638809e-05,
+      "loss": 2.6135,
+      "step": 241600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2292667307076534e-05,
+      "loss": 2.6242,
+      "step": 241700
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.2289465257764973e-05,
+      "loss": 2.6085,
+      "step": 241800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.228626320845341e-05,
+      "loss": 2.6124,
+      "step": 241900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.228306115914185e-05,
+      "loss": 2.6433,
+      "step": 242000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.5190274715423584,
+      "eval_runtime": 176.3523,
+      "eval_samples_per_second": 56.705,
+      "eval_steps_per_second": 3.544,
+      "step": 242000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.227985910983029e-05,
+      "loss": 2.6167,
+      "step": 242100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.227665706051873e-05,
+      "loss": 2.5979,
+      "step": 242200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.227345501120717e-05,
+      "loss": 2.6175,
+      "step": 242300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.227025296189562e-05,
+      "loss": 2.6009,
+      "step": 242400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.226705091258406e-05,
+      "loss": 2.5928,
+      "step": 242500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.22638488632725e-05,
+      "loss": 2.6029,
+      "step": 242600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.226064681396094e-05,
+      "loss": 2.6182,
+      "step": 242700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.225744476464938e-05,
+      "loss": 2.6158,
+      "step": 242800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.225424271533782e-05,
+      "loss": 2.6035,
+      "step": 242900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.225104066602626e-05,
+      "loss": 2.6024,
+      "step": 243000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5197348594665527,
+      "eval_runtime": 178.619,
+      "eval_samples_per_second": 55.985,
+      "eval_steps_per_second": 3.499,
+      "step": 243000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.22478386167147e-05,
+      "loss": 2.5982,
+      "step": 243100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2244636567403137e-05,
+      "loss": 2.6126,
+      "step": 243200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.224143451809158e-05,
+      "loss": 2.6022,
+      "step": 243300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.223823246878002e-05,
+      "loss": 2.6159,
+      "step": 243400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.223503041946846e-05,
+      "loss": 2.602,
+      "step": 243500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.22318283701569e-05,
+      "loss": 2.5958,
+      "step": 243600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.222862632084534e-05,
+      "loss": 2.5933,
+      "step": 243700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.222542427153378e-05,
+      "loss": 2.6133,
+      "step": 243800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 2.5952,
+      "step": 243900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.221902017291067e-05,
+      "loss": 2.6234,
+      "step": 244000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5188870429992676,
+      "eval_runtime": 176.9188,
+      "eval_samples_per_second": 56.523,
+      "eval_steps_per_second": 3.533,
+      "step": 244000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.221581812359911e-05,
+      "loss": 2.5892,
+      "step": 244100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.221261607428755e-05,
+      "loss": 2.6068,
+      "step": 244200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.220941402497599e-05,
+      "loss": 2.599,
+      "step": 244300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.220621197566443e-05,
+      "loss": 2.6054,
+      "step": 244400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.220300992635287e-05,
+      "loss": 2.6022,
+      "step": 244500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2199807877041306e-05,
+      "loss": 2.6111,
+      "step": 244600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.219660582772975e-05,
+      "loss": 2.5955,
+      "step": 244700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2193403778418186e-05,
+      "loss": 2.6128,
+      "step": 244800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.219020172910663e-05,
+      "loss": 2.5816,
+      "step": 244900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.218699967979507e-05,
+      "loss": 2.6044,
+      "step": 245000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.519922971725464,
+      "eval_runtime": 176.5307,
+      "eval_samples_per_second": 56.647,
+      "eval_steps_per_second": 3.54,
+      "step": 245000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.218379763048351e-05,
+      "loss": 2.5932,
+      "step": 245100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.218059558117195e-05,
+      "loss": 2.5801,
+      "step": 245200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.217739353186039e-05,
+      "loss": 2.6113,
+      "step": 245300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.217419148254884e-05,
+      "loss": 2.5939,
+      "step": 245400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.217098943323727e-05,
+      "loss": 2.606,
+      "step": 245500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.216778738392572e-05,
+      "loss": 2.6075,
+      "step": 245600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.216458533461416e-05,
+      "loss": 2.5771,
+      "step": 245700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.21613832853026e-05,
+      "loss": 2.6056,
+      "step": 245800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.215818123599104e-05,
+      "loss": 2.606,
+      "step": 245900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2154979186679476e-05,
+      "loss": 2.585,
+      "step": 246000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.521212577819824,
+      "eval_runtime": 174.8676,
+      "eval_samples_per_second": 57.186,
+      "eval_steps_per_second": 3.574,
+      "step": 246000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2151777137367916e-05,
+      "loss": 2.6218,
+      "step": 246100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2148575088056356e-05,
+      "loss": 2.6135,
+      "step": 246200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.21453730387448e-05,
+      "loss": 2.5745,
+      "step": 246300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2142170989433235e-05,
+      "loss": 2.5689,
+      "step": 246400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.213896894012168e-05,
+      "loss": 2.611,
+      "step": 246500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.213576689081012e-05,
+      "loss": 2.6038,
+      "step": 246600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.213256484149856e-05,
+      "loss": 2.6005,
+      "step": 246700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2129362792187e-05,
+      "loss": 2.5979,
+      "step": 246800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.212616074287544e-05,
+      "loss": 2.6079,
+      "step": 246900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.212295869356389e-05,
+      "loss": 2.6015,
+      "step": 247000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.519866466522217,
+      "eval_runtime": 177.7026,
+      "eval_samples_per_second": 56.274,
+      "eval_steps_per_second": 3.517,
+      "step": 247000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.211975664425232e-05,
+      "loss": 2.5989,
+      "step": 247100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.211655459494077e-05,
+      "loss": 2.6152,
+      "step": 247200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2113352545629207e-05,
+      "loss": 2.5921,
+      "step": 247300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2110150496317646e-05,
+      "loss": 2.586,
+      "step": 247400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2106948447006086e-05,
+      "loss": 2.6009,
+      "step": 247500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2103746397694526e-05,
+      "loss": 2.5924,
+      "step": 247600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.210054434838297e-05,
+      "loss": 2.5901,
+      "step": 247700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2097342299071405e-05,
+      "loss": 2.5757,
+      "step": 247800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.209414024975985e-05,
+      "loss": 2.5997,
+      "step": 247900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2090938200448285e-05,
+      "loss": 2.5898,
+      "step": 248000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.517677068710327,
+      "eval_runtime": 179.5331,
+      "eval_samples_per_second": 55.7,
+      "eval_steps_per_second": 3.481,
+      "step": 248000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.208773615113673e-05,
+      "loss": 2.6087,
+      "step": 248100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.208453410182517e-05,
+      "loss": 2.5959,
+      "step": 248200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.208133205251361e-05,
+      "loss": 2.5946,
+      "step": 248300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.207813000320205e-05,
+      "loss": 2.5904,
+      "step": 248400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.207492795389049e-05,
+      "loss": 2.6008,
+      "step": 248500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.207172590457894e-05,
+      "loss": 2.6257,
+      "step": 248600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.206852385526737e-05,
+      "loss": 2.5861,
+      "step": 248700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2065321805955816e-05,
+      "loss": 2.6015,
+      "step": 248800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2062119756644256e-05,
+      "loss": 2.6052,
+      "step": 248900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2058917707332696e-05,
+      "loss": 2.6083,
+      "step": 249000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5183045864105225,
+      "eval_runtime": 174.4514,
+      "eval_samples_per_second": 57.323,
+      "eval_steps_per_second": 3.583,
+      "step": 249000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2055715658021135e-05,
+      "loss": 2.5813,
+      "step": 249100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2052513608709575e-05,
+      "loss": 2.6023,
+      "step": 249200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.204931155939802e-05,
+      "loss": 2.6073,
+      "step": 249300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2046109510086455e-05,
+      "loss": 2.6034,
+      "step": 249400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.20429074607749e-05,
+      "loss": 2.5924,
+      "step": 249500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2039705411463334e-05,
+      "loss": 2.6071,
+      "step": 249600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.203650336215178e-05,
+      "loss": 2.6096,
+      "step": 249700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.203330131284022e-05,
+      "loss": 2.6003,
+      "step": 249800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.203009926352866e-05,
+      "loss": 2.5857,
+      "step": 249900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2026897214217107e-05,
+      "loss": 2.5914,
+      "step": 250000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5187132358551025,
+      "eval_runtime": 178.1664,
+      "eval_samples_per_second": 56.127,
+      "eval_steps_per_second": 3.508,
+      "step": 250000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.202369516490554e-05,
+      "loss": 2.5998,
+      "step": 250100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2020493115593986e-05,
+      "loss": 2.5964,
+      "step": 250200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.201729106628242e-05,
+      "loss": 2.6038,
+      "step": 250300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2014089016970866e-05,
+      "loss": 2.6056,
+      "step": 250400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2010886967659305e-05,
+      "loss": 2.6148,
+      "step": 250500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2007684918347745e-05,
+      "loss": 2.6028,
+      "step": 250600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2004482869036185e-05,
+      "loss": 2.6202,
+      "step": 250700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2001280819724624e-05,
+      "loss": 2.6139,
+      "step": 250800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.199807877041307e-05,
+      "loss": 2.5886,
+      "step": 250900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1994876721101504e-05,
+      "loss": 2.5908,
+      "step": 251000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5188255310058594,
+      "eval_runtime": 175.9361,
+      "eval_samples_per_second": 56.839,
+      "eval_steps_per_second": 3.552,
+      "step": 251000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.199167467178995e-05,
+      "loss": 2.6054,
+      "step": 251100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1988472622478383e-05,
+      "loss": 2.6235,
+      "step": 251200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.198527057316683e-05,
+      "loss": 2.6033,
+      "step": 251300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.198206852385527e-05,
+      "loss": 2.6026,
+      "step": 251400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.197886647454371e-05,
+      "loss": 2.5992,
+      "step": 251500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1975664425232156e-05,
+      "loss": 2.5972,
+      "step": 251600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.197246237592059e-05,
+      "loss": 2.6024,
+      "step": 251700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1969260326609035e-05,
+      "loss": 2.5899,
+      "step": 251800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.196605827729747e-05,
+      "loss": 2.5926,
+      "step": 251900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1962856227985915e-05,
+      "loss": 2.5775,
+      "step": 252000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5177624225616455,
+      "eval_runtime": 178.4144,
+      "eval_samples_per_second": 56.049,
+      "eval_steps_per_second": 3.503,
+      "step": 252000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1959654178674355e-05,
+      "loss": 2.6064,
+      "step": 252100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1956452129362794e-05,
+      "loss": 2.5807,
+      "step": 252200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1953250080051234e-05,
+      "loss": 2.576,
+      "step": 252300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1950048030739674e-05,
+      "loss": 2.6154,
+      "step": 252400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.194684598142812e-05,
+      "loss": 2.5894,
+      "step": 252500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.194364393211655e-05,
+      "loss": 2.5964,
+      "step": 252600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1940441882805e-05,
+      "loss": 2.5941,
+      "step": 252700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.193723983349343e-05,
+      "loss": 2.6035,
+      "step": 252800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.193403778418188e-05,
+      "loss": 2.5959,
+      "step": 252900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.193083573487032e-05,
+      "loss": 2.6098,
+      "step": 253000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.51733660697937,
+      "eval_runtime": 177.19,
+      "eval_samples_per_second": 56.437,
+      "eval_steps_per_second": 3.527,
+      "step": 253000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.192763368555876e-05,
+      "loss": 2.6199,
+      "step": 253100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1924431636247205e-05,
+      "loss": 2.5878,
+      "step": 253200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.192122958693564e-05,
+      "loss": 2.6154,
+      "step": 253300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1918027537624085e-05,
+      "loss": 2.6085,
+      "step": 253400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.191482548831252e-05,
+      "loss": 2.5945,
+      "step": 253500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1911623439000964e-05,
+      "loss": 2.5918,
+      "step": 253600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1908421389689404e-05,
+      "loss": 2.593,
+      "step": 253700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1905219340377844e-05,
+      "loss": 2.6045,
+      "step": 253800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1902017291066283e-05,
+      "loss": 2.582,
+      "step": 253900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.189881524175472e-05,
+      "loss": 2.5908,
+      "step": 254000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.517996072769165,
+      "eval_runtime": 179.2284,
+      "eval_samples_per_second": 55.795,
+      "eval_steps_per_second": 3.487,
+      "step": 254000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.189561319244317e-05,
+      "loss": 2.6161,
+      "step": 254100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.18924111431316e-05,
+      "loss": 2.5769,
+      "step": 254200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.188920909382005e-05,
+      "loss": 2.6037,
+      "step": 254300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.188600704450848e-05,
+      "loss": 2.5995,
+      "step": 254400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.188280499519693e-05,
+      "loss": 2.5857,
+      "step": 254500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.187960294588537e-05,
+      "loss": 2.5969,
+      "step": 254600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.187640089657381e-05,
+      "loss": 2.5733,
+      "step": 254700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1873198847262255e-05,
+      "loss": 2.6007,
+      "step": 254800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.186999679795069e-05,
+      "loss": 2.6058,
+      "step": 254900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1866794748639134e-05,
+      "loss": 2.6154,
+      "step": 255000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.517491102218628,
+      "eval_runtime": 176.7253,
+      "eval_samples_per_second": 56.585,
+      "eval_steps_per_second": 3.537,
+      "step": 255000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.186359269932757e-05,
+      "loss": 2.6082,
+      "step": 255100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1860390650016014e-05,
+      "loss": 2.6015,
+      "step": 255200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1857188600704453e-05,
+      "loss": 2.6023,
+      "step": 255300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.185398655139289e-05,
+      "loss": 2.5929,
+      "step": 255400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.185078450208133e-05,
+      "loss": 2.604,
+      "step": 255500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.184758245276977e-05,
+      "loss": 2.6106,
+      "step": 255600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.184438040345822e-05,
+      "loss": 2.5907,
+      "step": 255700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.184117835414665e-05,
+      "loss": 2.5968,
+      "step": 255800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.18379763048351e-05,
+      "loss": 2.59,
+      "step": 255900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.183477425552353e-05,
+      "loss": 2.6208,
+      "step": 256000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.516918182373047,
+      "eval_runtime": 173.799,
+      "eval_samples_per_second": 57.538,
+      "eval_steps_per_second": 3.596,
+      "step": 256000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.183157220621198e-05,
+      "loss": 2.5787,
+      "step": 256100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.182837015690042e-05,
+      "loss": 2.6033,
+      "step": 256200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.182516810758886e-05,
+      "loss": 2.5794,
+      "step": 256300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1821966058277304e-05,
+      "loss": 2.6046,
+      "step": 256400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.181876400896574e-05,
+      "loss": 2.6119,
+      "step": 256500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1815561959654184e-05,
+      "loss": 2.5908,
+      "step": 256600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.1812359910342617e-05,
+      "loss": 2.5921,
+      "step": 256700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.180915786103106e-05,
+      "loss": 2.6081,
+      "step": 256800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.18059558117195e-05,
+      "loss": 2.605,
+      "step": 256900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.180275376240794e-05,
+      "loss": 2.5885,
+      "step": 257000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5189480781555176,
+      "eval_runtime": 176.498,
+      "eval_samples_per_second": 56.658,
+      "eval_steps_per_second": 3.541,
+      "step": 257000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.179955171309638e-05,
+      "loss": 2.6032,
+      "step": 257100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.179634966378482e-05,
+      "loss": 2.5909,
+      "step": 257200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.179314761447327e-05,
+      "loss": 2.5831,
+      "step": 257300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.17899455651617e-05,
+      "loss": 2.6053,
+      "step": 257400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.178674351585015e-05,
+      "loss": 2.5832,
+      "step": 257500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.178354146653859e-05,
+      "loss": 2.5843,
+      "step": 257600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.178033941722703e-05,
+      "loss": 2.6138,
+      "step": 257700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.177713736791547e-05,
+      "loss": 2.5878,
+      "step": 257800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.177393531860391e-05,
+      "loss": 2.5953,
+      "step": 257900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1770733269292353e-05,
+      "loss": 2.5822,
+      "step": 258000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.5180604457855225,
+      "eval_runtime": 174.2586,
+      "eval_samples_per_second": 57.386,
+      "eval_steps_per_second": 3.587,
+      "step": 258000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1767531219980786e-05,
+      "loss": 2.5609,
+      "step": 258100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.176432917066923e-05,
+      "loss": 2.5931,
+      "step": 258200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1761127121357666e-05,
+      "loss": 2.5921,
+      "step": 258300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.175792507204611e-05,
+      "loss": 2.5857,
+      "step": 258400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.175472302273455e-05,
+      "loss": 2.5527,
+      "step": 258500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.175152097342299e-05,
+      "loss": 2.606,
+      "step": 258600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.174831892411143e-05,
+      "loss": 2.6038,
+      "step": 258700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.174511687479987e-05,
+      "loss": 2.5795,
+      "step": 258800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.174191482548832e-05,
+      "loss": 2.59,
+      "step": 258900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.173871277617675e-05,
+      "loss": 2.6116,
+      "step": 259000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.5164027214050293,
+      "eval_runtime": 177.8193,
+      "eval_samples_per_second": 56.237,
+      "eval_steps_per_second": 3.515,
+      "step": 259000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.17355107268652e-05,
+      "loss": 2.5925,
+      "step": 259100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.173230867755364e-05,
+      "loss": 2.5748,
+      "step": 259200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.172910662824208e-05,
+      "loss": 2.5917,
+      "step": 259300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1725904578930517e-05,
+      "loss": 2.5977,
+      "step": 259400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1722702529618956e-05,
+      "loss": 2.5957,
+      "step": 259500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.17195004803074e-05,
+      "loss": 2.599,
+      "step": 259600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1716298430995836e-05,
+      "loss": 2.571,
+      "step": 259700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.171309638168428e-05,
+      "loss": 2.5959,
+      "step": 259800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.170989433237272e-05,
+      "loss": 2.5897,
+      "step": 259900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.170669228306116e-05,
+      "loss": 2.5842,
+      "step": 260000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.515014410018921,
+      "eval_runtime": 177.7592,
+      "eval_samples_per_second": 56.256,
+      "eval_steps_per_second": 3.516,
+      "step": 260000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.17034902337496e-05,
+      "loss": 2.5836,
+      "step": 260100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.170028818443804e-05,
+      "loss": 2.5905,
+      "step": 260200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.169708613512648e-05,
+      "loss": 2.5985,
+      "step": 260300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.169388408581492e-05,
+      "loss": 2.5889,
+      "step": 260400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.169068203650337e-05,
+      "loss": 2.5939,
+      "step": 260500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.16874799871918e-05,
+      "loss": 2.5975,
+      "step": 260600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.168427793788025e-05,
+      "loss": 2.5703,
+      "step": 260700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1681075888568686e-05,
+      "loss": 2.61,
+      "step": 260800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1677873839257126e-05,
+      "loss": 2.5877,
+      "step": 260900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1674671789945566e-05,
+      "loss": 2.5998,
+      "step": 261000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.513521909713745,
+      "eval_runtime": 176.0418,
+      "eval_samples_per_second": 56.805,
+      "eval_steps_per_second": 3.55,
+      "step": 261000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1671469740634006e-05,
+      "loss": 2.5755,
+      "step": 261100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.166826769132245e-05,
+      "loss": 2.5838,
+      "step": 261200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1665065642010885e-05,
+      "loss": 2.5788,
+      "step": 261300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.166186359269933e-05,
+      "loss": 2.5935,
+      "step": 261400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.165866154338777e-05,
+      "loss": 2.5933,
+      "step": 261500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.165545949407621e-05,
+      "loss": 2.5783,
+      "step": 261600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.165225744476465e-05,
+      "loss": 2.5971,
+      "step": 261700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.164905539545309e-05,
+      "loss": 2.5905,
+      "step": 261800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.164585334614153e-05,
+      "loss": 2.5828,
+      "step": 261900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.164265129682997e-05,
+      "loss": 2.582,
+      "step": 262000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.5153019428253174,
+      "eval_runtime": 176.8144,
+      "eval_samples_per_second": 56.556,
+      "eval_steps_per_second": 3.535,
+      "step": 262000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1639449247518417e-05,
+      "loss": 2.6019,
+      "step": 262100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1636247198206856e-05,
+      "loss": 2.5596,
+      "step": 262200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1633045148895296e-05,
+      "loss": 2.6009,
+      "step": 262300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1629843099583736e-05,
+      "loss": 2.5817,
+      "step": 262400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1626641050272176e-05,
+      "loss": 2.5961,
+      "step": 262500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1623439000960615e-05,
+      "loss": 2.5779,
+      "step": 262600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1620236951649055e-05,
+      "loss": 2.5732,
+      "step": 262700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.16170349023375e-05,
+      "loss": 2.5881,
+      "step": 262800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.161383285302594e-05,
+      "loss": 2.5831,
+      "step": 262900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.161063080371438e-05,
+      "loss": 2.5916,
+      "step": 263000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.515854597091675,
+      "eval_runtime": 177.8685,
+      "eval_samples_per_second": 56.221,
+      "eval_steps_per_second": 3.514,
+      "step": 263000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.160742875440282e-05,
+      "loss": 2.5857,
+      "step": 263100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.160422670509126e-05,
+      "loss": 2.5812,
+      "step": 263200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.16010246557797e-05,
+      "loss": 2.5848,
+      "step": 263300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.159782260646814e-05,
+      "loss": 2.5676,
+      "step": 263400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.159462055715658e-05,
+      "loss": 2.5976,
+      "step": 263500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.159141850784502e-05,
+      "loss": 2.5659,
+      "step": 263600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1588216458533466e-05,
+      "loss": 2.5824,
+      "step": 263700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1585014409221906e-05,
+      "loss": 2.5917,
+      "step": 263800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1581812359910345e-05,
+      "loss": 2.5712,
+      "step": 263900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1578610310598785e-05,
+      "loss": 2.594,
+      "step": 264000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.515024185180664,
+      "eval_runtime": 177.996,
+      "eval_samples_per_second": 56.181,
+      "eval_steps_per_second": 3.511,
+      "step": 264000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1575408261287225e-05,
+      "loss": 2.5841,
+      "step": 264100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1572206211975665e-05,
+      "loss": 2.592,
+      "step": 264200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1569004162664104e-05,
+      "loss": 2.5938,
+      "step": 264300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.156580211335255e-05,
+      "loss": 2.5704,
+      "step": 264400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.156260006404099e-05,
+      "loss": 2.5664,
+      "step": 264500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.155939801472943e-05,
+      "loss": 2.576,
+      "step": 264600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.155619596541787e-05,
+      "loss": 2.5821,
+      "step": 264700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.155299391610631e-05,
+      "loss": 2.588,
+      "step": 264800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.154979186679475e-05,
+      "loss": 2.571,
+      "step": 264900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.154658981748319e-05,
+      "loss": 2.56,
+      "step": 265000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.5161280632019043,
+      "eval_runtime": 177.1926,
+      "eval_samples_per_second": 56.436,
+      "eval_steps_per_second": 3.527,
+      "step": 265000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.154338776817163e-05,
+      "loss": 2.5877,
+      "step": 265100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1540185718860076e-05,
+      "loss": 2.5833,
+      "step": 265200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1536983669548515e-05,
+      "loss": 2.5962,
+      "step": 265300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1533781620236955e-05,
+      "loss": 2.5718,
+      "step": 265400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1530579570925395e-05,
+      "loss": 2.5899,
+      "step": 265500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1527377521613835e-05,
+      "loss": 2.595,
+      "step": 265600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1524175472302274e-05,
+      "loss": 2.5843,
+      "step": 265700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1520973422990714e-05,
+      "loss": 2.5903,
+      "step": 265800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1517771373679154e-05,
+      "loss": 2.5776,
+      "step": 265900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.15145693243676e-05,
+      "loss": 2.5732,
+      "step": 266000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.515674114227295,
+      "eval_runtime": 176.8798,
+      "eval_samples_per_second": 56.536,
+      "eval_steps_per_second": 3.533,
+      "step": 266000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.151136727505604e-05,
+      "loss": 2.5623,
+      "step": 266100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.150816522574448e-05,
+      "loss": 2.5807,
+      "step": 266200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.150496317643292e-05,
+      "loss": 2.5973,
+      "step": 266300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.150176112712136e-05,
+      "loss": 2.5657,
+      "step": 266400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.14985590778098e-05,
+      "loss": 2.5731,
+      "step": 266500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.149535702849824e-05,
+      "loss": 2.5915,
+      "step": 266600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.149215497918668e-05,
+      "loss": 2.5775,
+      "step": 266700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1488952929875125e-05,
+      "loss": 2.5722,
+      "step": 266800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1485750880563565e-05,
+      "loss": 2.5907,
+      "step": 266900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1482548831252004e-05,
+      "loss": 2.5999,
+      "step": 267000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.5125646591186523,
+      "eval_runtime": 177.7706,
+      "eval_samples_per_second": 56.252,
+      "eval_steps_per_second": 3.516,
+      "step": 267000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1479346781940444e-05,
+      "loss": 2.5765,
+      "step": 267100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1476144732628884e-05,
+      "loss": 2.5961,
+      "step": 267200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1472942683317324e-05,
+      "loss": 2.5974,
+      "step": 267300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1469740634005763e-05,
+      "loss": 2.6007,
+      "step": 267400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.146653858469421e-05,
+      "loss": 2.5798,
+      "step": 267500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.146333653538265e-05,
+      "loss": 2.5943,
+      "step": 267600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.146013448607109e-05,
+      "loss": 2.5976,
+      "step": 267700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.145693243675953e-05,
+      "loss": 2.599,
+      "step": 267800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.145373038744797e-05,
+      "loss": 2.5967,
+      "step": 267900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.145052833813641e-05,
+      "loss": 2.5996,
+      "step": 268000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.512153148651123,
+      "eval_runtime": 175.9425,
+      "eval_samples_per_second": 56.837,
+      "eval_steps_per_second": 3.552,
+      "step": 268000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.144732628882485e-05,
+      "loss": 2.5637,
+      "step": 268100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.144412423951329e-05,
+      "loss": 2.6014,
+      "step": 268200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.144092219020173e-05,
+      "loss": 2.5843,
+      "step": 268300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1437720140890174e-05,
+      "loss": 2.5961,
+      "step": 268400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1434518091578614e-05,
+      "loss": 2.5788,
+      "step": 268500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1431316042267054e-05,
+      "loss": 2.5749,
+      "step": 268600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1428113992955494e-05,
+      "loss": 2.5783,
+      "step": 268700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.142491194364393e-05,
+      "loss": 2.6043,
+      "step": 268800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.142170989433237e-05,
+      "loss": 2.5832,
+      "step": 268900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.141850784502081e-05,
+      "loss": 2.5844,
+      "step": 269000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.512434482574463,
+      "eval_runtime": 177.2599,
+      "eval_samples_per_second": 56.414,
+      "eval_steps_per_second": 3.526,
+      "step": 269000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.141530579570926e-05,
+      "loss": 2.6106,
+      "step": 269100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.14121037463977e-05,
+      "loss": 2.6047,
+      "step": 269200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.140890169708614e-05,
+      "loss": 2.5827,
+      "step": 269300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.140569964777458e-05,
+      "loss": 2.5831,
+      "step": 269400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.140249759846302e-05,
+      "loss": 2.5736,
+      "step": 269500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.139929554915146e-05,
+      "loss": 2.5827,
+      "step": 269600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.13960934998399e-05,
+      "loss": 2.5672,
+      "step": 269700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1392891450528344e-05,
+      "loss": 2.5707,
+      "step": 269800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.138968940121678e-05,
+      "loss": 2.5966,
+      "step": 269900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1386487351905224e-05,
+      "loss": 2.5745,
+      "step": 270000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.5134899616241455,
+      "eval_runtime": 180.3244,
+      "eval_samples_per_second": 55.456,
+      "eval_steps_per_second": 3.466,
+      "step": 270000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1383285302593663e-05,
+      "loss": 2.5802,
+      "step": 270100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.13800832532821e-05,
+      "loss": 2.5698,
+      "step": 270200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.137688120397054e-05,
+      "loss": 2.5655,
+      "step": 270300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.137367915465898e-05,
+      "loss": 2.5857,
+      "step": 270400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.137047710534742e-05,
+      "loss": 2.5985,
+      "step": 270500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.136727505603586e-05,
+      "loss": 2.5936,
+      "step": 270600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.136407300672431e-05,
+      "loss": 2.5917,
+      "step": 270700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.136087095741275e-05,
+      "loss": 2.5781,
+      "step": 270800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.135766890810119e-05,
+      "loss": 2.5733,
+      "step": 270900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.135446685878963e-05,
+      "loss": 2.5809,
+      "step": 271000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.5146307945251465,
+      "eval_runtime": 177.0724,
+      "eval_samples_per_second": 56.474,
+      "eval_steps_per_second": 3.53,
+      "step": 271000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.135126480947807e-05,
+      "loss": 2.5765,
+      "step": 271100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.134806276016651e-05,
+      "loss": 2.585,
+      "step": 271200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.134486071085495e-05,
+      "loss": 2.5658,
+      "step": 271300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1341658661543394e-05,
+      "loss": 2.574,
+      "step": 271400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1338456612231827e-05,
+      "loss": 2.5854,
+      "step": 271500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.133525456292027e-05,
+      "loss": 2.5767,
+      "step": 271600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.133205251360871e-05,
+      "loss": 2.5914,
+      "step": 271700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.132885046429715e-05,
+      "loss": 2.5716,
+      "step": 271800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.132564841498559e-05,
+      "loss": 2.579,
+      "step": 271900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.132244636567403e-05,
+      "loss": 2.5993,
+      "step": 272000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.513415575027466,
+      "eval_runtime": 176.3837,
+      "eval_samples_per_second": 56.695,
+      "eval_steps_per_second": 3.543,
+      "step": 272000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.131924431636248e-05,
+      "loss": 2.5811,
+      "step": 272100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.131604226705091e-05,
+      "loss": 2.5801,
+      "step": 272200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.131284021773936e-05,
+      "loss": 2.5834,
+      "step": 272300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.13096381684278e-05,
+      "loss": 2.6039,
+      "step": 272400
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.130643611911624e-05,
+      "loss": 2.5798,
+      "step": 272500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.130323406980468e-05,
+      "loss": 2.5923,
+      "step": 272600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.130003202049312e-05,
+      "loss": 2.585,
+      "step": 272700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1296829971181564e-05,
+      "loss": 2.5654,
+      "step": 272800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1293627921869996e-05,
+      "loss": 2.6044,
+      "step": 272900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.129042587255844e-05,
+      "loss": 2.5672,
+      "step": 273000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.514824151992798,
+      "eval_runtime": 176.7123,
+      "eval_samples_per_second": 56.589,
+      "eval_steps_per_second": 3.537,
+      "step": 273000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.1287223823246876e-05,
+      "loss": 2.5791,
+      "step": 273100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.128402177393532e-05,
+      "loss": 2.5671,
+      "step": 273200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.128081972462376e-05,
+      "loss": 2.5685,
+      "step": 273300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.12776176753122e-05,
+      "loss": 2.5751,
+      "step": 273400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.127441562600064e-05,
+      "loss": 2.542,
+      "step": 273500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.127121357668908e-05,
+      "loss": 2.5675,
+      "step": 273600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.126801152737753e-05,
+      "loss": 2.5709,
+      "step": 273700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.126480947806596e-05,
+      "loss": 2.5776,
+      "step": 273800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.126160742875441e-05,
+      "loss": 2.5686,
+      "step": 273900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.125840537944285e-05,
+      "loss": 2.5814,
+      "step": 274000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.512730836868286,
+      "eval_runtime": 178.1525,
+      "eval_samples_per_second": 56.132,
+      "eval_steps_per_second": 3.508,
+      "step": 274000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.125520333013129e-05,
+      "loss": 2.5885,
+      "step": 274100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.125200128081973e-05,
+      "loss": 2.5774,
+      "step": 274200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1248799231508166e-05,
+      "loss": 2.5668,
+      "step": 274300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.124559718219661e-05,
+      "loss": 2.5889,
+      "step": 274400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1242395132885046e-05,
+      "loss": 2.5962,
+      "step": 274500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.123919308357349e-05,
+      "loss": 2.5924,
+      "step": 274600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1235991034261925e-05,
+      "loss": 2.5818,
+      "step": 274700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.123278898495037e-05,
+      "loss": 2.5815,
+      "step": 274800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.122958693563881e-05,
+      "loss": 2.5683,
+      "step": 274900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.122638488632725e-05,
+      "loss": 2.5661,
+      "step": 275000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.510558843612671,
+      "eval_runtime": 178.7649,
+      "eval_samples_per_second": 55.939,
+      "eval_steps_per_second": 3.496,
+      "step": 275000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.12231828370157e-05,
+      "loss": 2.5731,
+      "step": 275100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.121998078770413e-05,
+      "loss": 2.56,
+      "step": 275200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.121677873839258e-05,
+      "loss": 2.5673,
+      "step": 275300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.121357668908101e-05,
+      "loss": 2.5788,
+      "step": 275400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.121037463976946e-05,
+      "loss": 2.5721,
+      "step": 275500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.120717259045789e-05,
+      "loss": 2.5839,
+      "step": 275600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1203970541146336e-05,
+      "loss": 2.576,
+      "step": 275700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1200768491834776e-05,
+      "loss": 2.5467,
+      "step": 275800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1197566442523216e-05,
+      "loss": 2.5834,
+      "step": 275900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.119436439321166e-05,
+      "loss": 2.5884,
+      "step": 276000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5124151706695557,
+      "eval_runtime": 174.6576,
+      "eval_samples_per_second": 57.255,
+      "eval_steps_per_second": 3.578,
+      "step": 276000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1191162343900095e-05,
+      "loss": 2.5852,
+      "step": 276100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.118796029458854e-05,
+      "loss": 2.5807,
+      "step": 276200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1184758245276975e-05,
+      "loss": 2.5777,
+      "step": 276300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.118155619596542e-05,
+      "loss": 2.5618,
+      "step": 276400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.117835414665386e-05,
+      "loss": 2.5635,
+      "step": 276500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.11751520973423e-05,
+      "loss": 2.5857,
+      "step": 276600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.117195004803075e-05,
+      "loss": 2.5788,
+      "step": 276700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.116874799871918e-05,
+      "loss": 2.5738,
+      "step": 276800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.116554594940763e-05,
+      "loss": 2.5889,
+      "step": 276900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.116234390009606e-05,
+      "loss": 2.5654,
+      "step": 277000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5116689205169678,
+      "eval_runtime": 176.8925,
+      "eval_samples_per_second": 56.532,
+      "eval_steps_per_second": 3.533,
+      "step": 277000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1159141850784506e-05,
+      "loss": 2.5913,
+      "step": 277100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.115593980147294e-05,
+      "loss": 2.5725,
+      "step": 277200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1152737752161386e-05,
+      "loss": 2.5543,
+      "step": 277300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1149535702849825e-05,
+      "loss": 2.5497,
+      "step": 277400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1146333653538265e-05,
+      "loss": 2.5553,
+      "step": 277500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.114313160422671e-05,
+      "loss": 2.5875,
+      "step": 277600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1139929554915145e-05,
+      "loss": 2.5723,
+      "step": 277700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.113672750560359e-05,
+      "loss": 2.5872,
+      "step": 277800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1133525456292024e-05,
+      "loss": 2.5567,
+      "step": 277900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.113032340698047e-05,
+      "loss": 2.5727,
+      "step": 278000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.511739492416382,
+      "eval_runtime": 174.9948,
+      "eval_samples_per_second": 57.145,
+      "eval_steps_per_second": 3.572,
+      "step": 278000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.112712135766891e-05,
+      "loss": 2.5517,
+      "step": 278100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.112391930835735e-05,
+      "loss": 2.5698,
+      "step": 278200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1120717259045797e-05,
+      "loss": 2.5745,
+      "step": 278300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.111751520973423e-05,
+      "loss": 2.5706,
+      "step": 278400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1114313160422676e-05,
+      "loss": 2.5668,
+      "step": 278500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 2.5859,
+      "step": 278600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1107909061799556e-05,
+      "loss": 2.5705,
+      "step": 278700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.110470701248799e-05,
+      "loss": 2.5645,
+      "step": 278800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1101504963176435e-05,
+      "loss": 2.5714,
+      "step": 278900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1098302913864875e-05,
+      "loss": 2.5648,
+      "step": 279000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.511039972305298,
+      "eval_runtime": 177.9542,
+      "eval_samples_per_second": 56.194,
+      "eval_steps_per_second": 3.512,
+      "step": 279000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1095100864553314e-05,
+      "loss": 2.5803,
+      "step": 279100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.109189881524176e-05,
+      "loss": 2.5774,
+      "step": 279200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1088696765930194e-05,
+      "loss": 2.5557,
+      "step": 279300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.108549471661864e-05,
+      "loss": 2.5804,
+      "step": 279400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1082292667307073e-05,
+      "loss": 2.5812,
+      "step": 279500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.107909061799552e-05,
+      "loss": 2.5659,
+      "step": 279600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.107588856868396e-05,
+      "loss": 2.5777,
+      "step": 279700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.10726865193724e-05,
+      "loss": 2.5677,
+      "step": 279800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1069484470060846e-05,
+      "loss": 2.5654,
+      "step": 279900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.106628242074928e-05,
+      "loss": 2.5633,
+      "step": 280000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5125532150268555,
+      "eval_runtime": 173.5232,
+      "eval_samples_per_second": 57.629,
+      "eval_steps_per_second": 3.602,
+      "step": 280000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1063080371437725e-05,
+      "loss": 2.5804,
+      "step": 280100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.105987832212616e-05,
+      "loss": 2.55,
+      "step": 280200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1056676272814605e-05,
+      "loss": 2.5984,
+      "step": 280300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.105347422350304e-05,
+      "loss": 2.5503,
+      "step": 280400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1050272174191484e-05,
+      "loss": 2.5609,
+      "step": 280500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1047070124879924e-05,
+      "loss": 2.5903,
+      "step": 280600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1043868075568364e-05,
+      "loss": 2.5634,
+      "step": 280700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.104066602625681e-05,
+      "loss": 2.5981,
+      "step": 280800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.103746397694524e-05,
+      "loss": 2.5627,
+      "step": 280900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.103426192763369e-05,
+      "loss": 2.5925,
+      "step": 281000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5105316638946533,
+      "eval_runtime": 175.7903,
+      "eval_samples_per_second": 56.886,
+      "eval_steps_per_second": 3.555,
+      "step": 281000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.103105987832212e-05,
+      "loss": 2.5634,
+      "step": 281100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.102785782901057e-05,
+      "loss": 2.5757,
+      "step": 281200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.102465577969901e-05,
+      "loss": 2.5716,
+      "step": 281300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.102145373038745e-05,
+      "loss": 2.5856,
+      "step": 281400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1018251681075895e-05,
+      "loss": 2.5613,
+      "step": 281500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.101504963176433e-05,
+      "loss": 2.5924,
+      "step": 281600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1011847582452775e-05,
+      "loss": 2.5446,
+      "step": 281700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.100864553314121e-05,
+      "loss": 2.5892,
+      "step": 281800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1005443483829654e-05,
+      "loss": 2.5723,
+      "step": 281900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1002241434518094e-05,
+      "loss": 2.581,
+      "step": 282000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5110504627227783,
+      "eval_runtime": 176.3019,
+      "eval_samples_per_second": 56.721,
+      "eval_steps_per_second": 3.545,
+      "step": 282000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0999039385206534e-05,
+      "loss": 2.5647,
+      "step": 282100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0995837335894973e-05,
+      "loss": 2.5775,
+      "step": 282200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.099263528658341e-05,
+      "loss": 2.5806,
+      "step": 282300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.098943323727186e-05,
+      "loss": 2.58,
+      "step": 282400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.098623118796029e-05,
+      "loss": 2.5629,
+      "step": 282500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.098302913864874e-05,
+      "loss": 2.5716,
+      "step": 282600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.097982708933718e-05,
+      "loss": 2.5695,
+      "step": 282700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.097662504002562e-05,
+      "loss": 2.5563,
+      "step": 282800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.097342299071406e-05,
+      "loss": 2.5978,
+      "step": 282900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.09702209414025e-05,
+      "loss": 2.5856,
+      "step": 283000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.511104106903076,
+      "eval_runtime": 179.631,
+      "eval_samples_per_second": 55.67,
+      "eval_steps_per_second": 3.479,
+      "step": 283000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0967018892090945e-05,
+      "loss": 2.5935,
+      "step": 283100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.096381684277938e-05,
+      "loss": 2.5626,
+      "step": 283200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0960614793467824e-05,
+      "loss": 2.5588,
+      "step": 283300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.095741274415626e-05,
+      "loss": 2.5614,
+      "step": 283400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0954210694844704e-05,
+      "loss": 2.5808,
+      "step": 283500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.095100864553314e-05,
+      "loss": 2.566,
+      "step": 283600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.094780659622158e-05,
+      "loss": 2.5677,
+      "step": 283700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.094460454691002e-05,
+      "loss": 2.5766,
+      "step": 283800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.094140249759846e-05,
+      "loss": 2.5525,
+      "step": 283900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.093820044828691e-05,
+      "loss": 2.5671,
+      "step": 284000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.509161949157715,
+      "eval_runtime": 176.3943,
+      "eval_samples_per_second": 56.691,
+      "eval_steps_per_second": 3.543,
+      "step": 284000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.093499839897534e-05,
+      "loss": 2.5692,
+      "step": 284100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.093179634966379e-05,
+      "loss": 2.5572,
+      "step": 284200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.092859430035223e-05,
+      "loss": 2.5831,
+      "step": 284300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.092539225104067e-05,
+      "loss": 2.5719,
+      "step": 284400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.092219020172911e-05,
+      "loss": 2.5671,
+      "step": 284500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.091898815241755e-05,
+      "loss": 2.5802,
+      "step": 284600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0915786103105994e-05,
+      "loss": 2.5717,
+      "step": 284700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.091258405379443e-05,
+      "loss": 2.5834,
+      "step": 284800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0909382004482874e-05,
+      "loss": 2.5793,
+      "step": 284900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.090617995517131e-05,
+      "loss": 2.5728,
+      "step": 285000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5098185539245605,
+      "eval_runtime": 173.9083,
+      "eval_samples_per_second": 57.502,
+      "eval_steps_per_second": 3.594,
+      "step": 285000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.090297790585975e-05,
+      "loss": 2.5591,
+      "step": 285100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.089977585654819e-05,
+      "loss": 2.5807,
+      "step": 285200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.089657380723663e-05,
+      "loss": 2.5771,
+      "step": 285300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.089337175792507e-05,
+      "loss": 2.5801,
+      "step": 285400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.089016970861351e-05,
+      "loss": 2.5605,
+      "step": 285500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.088696765930196e-05,
+      "loss": 2.5708,
+      "step": 285600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.088376560999039e-05,
+      "loss": 2.5581,
+      "step": 285700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.088056356067884e-05,
+      "loss": 2.5597,
+      "step": 285800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.087736151136728e-05,
+      "loss": 2.553,
+      "step": 285900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.087415946205572e-05,
+      "loss": 2.5626,
+      "step": 286000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.51062273979187,
+      "eval_runtime": 176.5074,
+      "eval_samples_per_second": 56.655,
+      "eval_steps_per_second": 3.541,
+      "step": 286000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.087095741274416e-05,
+      "loss": 2.5574,
+      "step": 286100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.08677553634326e-05,
+      "loss": 2.562,
+      "step": 286200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0864553314121043e-05,
+      "loss": 2.5496,
+      "step": 286300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0861351264809476e-05,
+      "loss": 2.5846,
+      "step": 286400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.085814921549792e-05,
+      "loss": 2.5862,
+      "step": 286500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.085494716618636e-05,
+      "loss": 2.5518,
+      "step": 286600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.08517451168748e-05,
+      "loss": 2.5733,
+      "step": 286700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.084854306756324e-05,
+      "loss": 2.5678,
+      "step": 286800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.084534101825168e-05,
+      "loss": 2.5802,
+      "step": 286900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.084213896894012e-05,
+      "loss": 2.5513,
+      "step": 287000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5107758045196533,
+      "eval_runtime": 175.9187,
+      "eval_samples_per_second": 56.844,
+      "eval_steps_per_second": 3.553,
+      "step": 287000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.083893691962856e-05,
+      "loss": 2.5619,
+      "step": 287100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.083573487031701e-05,
+      "loss": 2.5645,
+      "step": 287200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.083253282100545e-05,
+      "loss": 2.5653,
+      "step": 287300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.082933077169389e-05,
+      "loss": 2.5906,
+      "step": 287400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.082612872238233e-05,
+      "loss": 2.5605,
+      "step": 287500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.082292667307077e-05,
+      "loss": 2.5506,
+      "step": 287600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0819724623759207e-05,
+      "loss": 2.5645,
+      "step": 287700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0816522574447646e-05,
+      "loss": 2.5771,
+      "step": 287800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.081332052513609e-05,
+      "loss": 2.5784,
+      "step": 287900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.0810118475824526e-05,
+      "loss": 2.5557,
+      "step": 288000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.5089871883392334,
+      "eval_runtime": 178.3368,
+      "eval_samples_per_second": 56.074,
+      "eval_steps_per_second": 3.505,
+      "step": 288000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.080691642651297e-05,
+      "loss": 2.5438,
+      "step": 288100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.080371437720141e-05,
+      "loss": 2.5454,
+      "step": 288200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.080051232788985e-05,
+      "loss": 2.579,
+      "step": 288300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.079731027857829e-05,
+      "loss": 2.5547,
+      "step": 288400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.079410822926673e-05,
+      "loss": 2.5641,
+      "step": 288500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.079090617995517e-05,
+      "loss": 2.5702,
+      "step": 288600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.078770413064361e-05,
+      "loss": 2.5474,
+      "step": 288700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.078450208133206e-05,
+      "loss": 2.5632,
+      "step": 288800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.07813000320205e-05,
+      "loss": 2.5475,
+      "step": 288900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.077809798270894e-05,
+      "loss": 2.571,
+      "step": 289000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.510617971420288,
+      "eval_runtime": 177.734,
+      "eval_samples_per_second": 56.264,
+      "eval_steps_per_second": 3.516,
+      "step": 289000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0774895933397376e-05,
+      "loss": 2.5685,
+      "step": 289100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0771693884085816e-05,
+      "loss": 2.5794,
+      "step": 289200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0768491834774256e-05,
+      "loss": 2.5354,
+      "step": 289300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0765289785462696e-05,
+      "loss": 2.5567,
+      "step": 289400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.076208773615114e-05,
+      "loss": 2.5689,
+      "step": 289500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.075888568683958e-05,
+      "loss": 2.5718,
+      "step": 289600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.075568363752802e-05,
+      "loss": 2.5576,
+      "step": 289700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.075248158821646e-05,
+      "loss": 2.5523,
+      "step": 289800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.07492795389049e-05,
+      "loss": 2.5517,
+      "step": 289900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.074607748959334e-05,
+      "loss": 2.5623,
+      "step": 290000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.508697748184204,
+      "eval_runtime": 176.5437,
+      "eval_samples_per_second": 56.643,
+      "eval_steps_per_second": 3.54,
+      "step": 290000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.074287544028178e-05,
+      "loss": 2.5643,
+      "step": 290100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.073967339097022e-05,
+      "loss": 2.5692,
+      "step": 290200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.073647134165866e-05,
+      "loss": 2.5541,
+      "step": 290300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0733269292347107e-05,
+      "loss": 2.5701,
+      "step": 290400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0730067243035546e-05,
+      "loss": 2.5417,
+      "step": 290500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0726865193723986e-05,
+      "loss": 2.5643,
+      "step": 290600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0723663144412426e-05,
+      "loss": 2.5483,
+      "step": 290700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0720461095100866e-05,
+      "loss": 2.5724,
+      "step": 290800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0717259045789305e-05,
+      "loss": 2.5341,
+      "step": 290900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0714056996477745e-05,
+      "loss": 2.5673,
+      "step": 291000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.508972644805908,
+      "eval_runtime": 177.4854,
+      "eval_samples_per_second": 56.343,
+      "eval_steps_per_second": 3.521,
+      "step": 291000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.071085494716619e-05,
+      "loss": 2.5623,
+      "step": 291100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.070765289785463e-05,
+      "loss": 2.578,
+      "step": 291200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.070445084854307e-05,
+      "loss": 2.5711,
+      "step": 291300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.070124879923151e-05,
+      "loss": 2.5647,
+      "step": 291400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.069804674991995e-05,
+      "loss": 2.5425,
+      "step": 291500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.069484470060839e-05,
+      "loss": 2.5546,
+      "step": 291600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.069164265129683e-05,
+      "loss": 2.5606,
+      "step": 291700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.068844060198527e-05,
+      "loss": 2.5569,
+      "step": 291800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0685238552673716e-05,
+      "loss": 2.5551,
+      "step": 291900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0682036503362156e-05,
+      "loss": 2.5544,
+      "step": 292000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.508294105529785,
+      "eval_runtime": 174.7106,
+      "eval_samples_per_second": 57.238,
+      "eval_steps_per_second": 3.577,
+      "step": 292000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0678834454050596e-05,
+      "loss": 2.5827,
+      "step": 292100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0675632404739035e-05,
+      "loss": 2.5554,
+      "step": 292200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0672430355427475e-05,
+      "loss": 2.5515,
+      "step": 292300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0669228306115915e-05,
+      "loss": 2.5602,
+      "step": 292400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0666026256804355e-05,
+      "loss": 2.562,
+      "step": 292500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.06628242074928e-05,
+      "loss": 2.5807,
+      "step": 292600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.065962215818124e-05,
+      "loss": 2.5839,
+      "step": 292700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.065642010886968e-05,
+      "loss": 2.548,
+      "step": 292800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.065321805955812e-05,
+      "loss": 2.5535,
+      "step": 292900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.065001601024656e-05,
+      "loss": 2.5625,
+      "step": 293000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.509491205215454,
+      "eval_runtime": 178.2991,
+      "eval_samples_per_second": 56.086,
+      "eval_steps_per_second": 3.505,
+      "step": 293000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0646813960935e-05,
+      "loss": 2.5633,
+      "step": 293100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.064361191162344e-05,
+      "loss": 2.5533,
+      "step": 293200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.064040986231188e-05,
+      "loss": 2.541,
+      "step": 293300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.063720781300032e-05,
+      "loss": 2.5639,
+      "step": 293400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0634005763688766e-05,
+      "loss": 2.547,
+      "step": 293500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0630803714377205e-05,
+      "loss": 2.5464,
+      "step": 293600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0627601665065645e-05,
+      "loss": 2.5627,
+      "step": 293700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0624399615754085e-05,
+      "loss": 2.5713,
+      "step": 293800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0621197566442525e-05,
+      "loss": 2.5626,
+      "step": 293900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0617995517130964e-05,
+      "loss": 2.5665,
+      "step": 294000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5079572200775146,
+      "eval_runtime": 174.8545,
+      "eval_samples_per_second": 57.19,
+      "eval_steps_per_second": 3.574,
+      "step": 294000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0614793467819404e-05,
+      "loss": 2.5639,
+      "step": 294100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.061159141850785e-05,
+      "loss": 2.5535,
+      "step": 294200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0608389369196284e-05,
+      "loss": 2.5441,
+      "step": 294300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.060518731988473e-05,
+      "loss": 2.5535,
+      "step": 294400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.060198527057317e-05,
+      "loss": 2.5606,
+      "step": 294500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.059878322126161e-05,
+      "loss": 2.5639,
+      "step": 294600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.059558117195005e-05,
+      "loss": 2.5637,
+      "step": 294700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.059237912263849e-05,
+      "loss": 2.5466,
+      "step": 294800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0589177073326935e-05,
+      "loss": 2.5829,
+      "step": 294900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.058597502401537e-05,
+      "loss": 2.5556,
+      "step": 295000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.509488344192505,
+      "eval_runtime": 178.3018,
+      "eval_samples_per_second": 56.085,
+      "eval_steps_per_second": 3.505,
+      "step": 295000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0582772974703815e-05,
+      "loss": 2.5414,
+      "step": 295100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0579570925392255e-05,
+      "loss": 2.5758,
+      "step": 295200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0576368876080694e-05,
+      "loss": 2.5638,
+      "step": 295300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0573166826769134e-05,
+      "loss": 2.5582,
+      "step": 295400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0569964777457574e-05,
+      "loss": 2.5373,
+      "step": 295500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0566762728146014e-05,
+      "loss": 2.5559,
+      "step": 295600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0563560678834453e-05,
+      "loss": 2.575,
+      "step": 295700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.05603586295229e-05,
+      "loss": 2.5523,
+      "step": 295800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.055715658021133e-05,
+      "loss": 2.5523,
+      "step": 295900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.055395453089978e-05,
+      "loss": 2.5525,
+      "step": 296000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5082788467407227,
+      "eval_runtime": 176.6723,
+      "eval_samples_per_second": 56.602,
+      "eval_steps_per_second": 3.538,
+      "step": 296000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.055075248158822e-05,
+      "loss": 2.5737,
+      "step": 296100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.054755043227666e-05,
+      "loss": 2.5611,
+      "step": 296200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.05443483829651e-05,
+      "loss": 2.5676,
+      "step": 296300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.054114633365354e-05,
+      "loss": 2.5752,
+      "step": 296400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0537944284341985e-05,
+      "loss": 2.5637,
+      "step": 296500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.053474223503042e-05,
+      "loss": 2.5388,
+      "step": 296600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0531540185718864e-05,
+      "loss": 2.5566,
+      "step": 296700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0528338136407304e-05,
+      "loss": 2.534,
+      "step": 296800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0525136087095744e-05,
+      "loss": 2.5518,
+      "step": 296900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0521934037784184e-05,
+      "loss": 2.5801,
+      "step": 297000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5086190700531006,
+      "eval_runtime": 174.1216,
+      "eval_samples_per_second": 57.431,
+      "eval_steps_per_second": 3.589,
+      "step": 297000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.051873198847262e-05,
+      "loss": 2.546,
+      "step": 297100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.051552993916107e-05,
+      "loss": 2.552,
+      "step": 297200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.05123278898495e-05,
+      "loss": 2.5856,
+      "step": 297300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.050912584053795e-05,
+      "loss": 2.5604,
+      "step": 297400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.050592379122638e-05,
+      "loss": 2.57,
+      "step": 297500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.050272174191483e-05,
+      "loss": 2.5799,
+      "step": 297600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.049951969260327e-05,
+      "loss": 2.5632,
+      "step": 297700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.049631764329171e-05,
+      "loss": 2.5524,
+      "step": 297800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.049311559398015e-05,
+      "loss": 2.5868,
+      "step": 297900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.048991354466859e-05,
+      "loss": 2.5558,
+      "step": 298000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5086541175842285,
+      "eval_runtime": 176.9276,
+      "eval_samples_per_second": 56.52,
+      "eval_steps_per_second": 3.533,
+      "step": 298000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0486711495357034e-05,
+      "loss": 2.5616,
+      "step": 298100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.048350944604547e-05,
+      "loss": 2.5705,
+      "step": 298200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0480307396733914e-05,
+      "loss": 2.5497,
+      "step": 298300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0477105347422353e-05,
+      "loss": 2.5514,
+      "step": 298400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.047390329811079e-05,
+      "loss": 2.5565,
+      "step": 298500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.047070124879923e-05,
+      "loss": 2.5641,
+      "step": 298600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.046749919948767e-05,
+      "loss": 2.569,
+      "step": 298700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.046429715017612e-05,
+      "loss": 2.5735,
+      "step": 298800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.046109510086455e-05,
+      "loss": 2.5601,
+      "step": 298900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0457893051553e-05,
+      "loss": 2.5677,
+      "step": 299000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5073604583740234,
+      "eval_runtime": 173.7708,
+      "eval_samples_per_second": 57.547,
+      "eval_steps_per_second": 3.597,
+      "step": 299000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.045469100224143e-05,
+      "loss": 2.5402,
+      "step": 299100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.045148895292988e-05,
+      "loss": 2.551,
+      "step": 299200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.044828690361832e-05,
+      "loss": 2.5308,
+      "step": 299300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.044508485430676e-05,
+      "loss": 2.535,
+      "step": 299400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0441882804995204e-05,
+      "loss": 2.5503,
+      "step": 299500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.043868075568364e-05,
+      "loss": 2.5448,
+      "step": 299600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0435478706372084e-05,
+      "loss": 2.5586,
+      "step": 299700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0432276657060517e-05,
+      "loss": 2.5536,
+      "step": 299800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.042907460774896e-05,
+      "loss": 2.5617,
+      "step": 299900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.04258725584374e-05,
+      "loss": 2.5559,
+      "step": 300000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5065934658050537,
+      "eval_runtime": 177.8009,
+      "eval_samples_per_second": 56.243,
+      "eval_steps_per_second": 3.515,
+      "step": 300000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.042267050912584e-05,
+      "loss": 2.5646,
+      "step": 300100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.041946845981428e-05,
+      "loss": 2.573,
+      "step": 300200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.041626641050272e-05,
+      "loss": 2.5649,
+      "step": 300300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.041306436119117e-05,
+      "loss": 2.5522,
+      "step": 300400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.04098623118796e-05,
+      "loss": 2.5513,
+      "step": 300500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.040666026256805e-05,
+      "loss": 2.5763,
+      "step": 300600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.040345821325648e-05,
+      "loss": 2.5712,
+      "step": 300700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.040025616394493e-05,
+      "loss": 2.5657,
+      "step": 300800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.039705411463337e-05,
+      "loss": 2.5598,
+      "step": 300900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.039385206532181e-05,
+      "loss": 2.5637,
+      "step": 301000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5084407329559326,
+      "eval_runtime": 177.1384,
+      "eval_samples_per_second": 56.453,
+      "eval_steps_per_second": 3.528,
+      "step": 301000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0390650016010254e-05,
+      "loss": 2.5361,
+      "step": 301100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0387447966698686e-05,
+      "loss": 2.556,
+      "step": 301200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.038424591738713e-05,
+      "loss": 2.5632,
+      "step": 301300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0381043868075566e-05,
+      "loss": 2.5759,
+      "step": 301400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.037784181876401e-05,
+      "loss": 2.5651,
+      "step": 301500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.037463976945245e-05,
+      "loss": 2.5534,
+      "step": 301600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.037143772014089e-05,
+      "loss": 2.5505,
+      "step": 301700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.036823567082934e-05,
+      "loss": 2.5649,
+      "step": 301800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.036503362151777e-05,
+      "loss": 2.5579,
+      "step": 301900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.036183157220622e-05,
+      "loss": 2.5369,
+      "step": 302000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.508272171020508,
+      "eval_runtime": 173.6825,
+      "eval_samples_per_second": 57.576,
+      "eval_steps_per_second": 3.599,
+      "step": 302000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.035862952289465e-05,
+      "loss": 2.5736,
+      "step": 302100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.03554274735831e-05,
+      "loss": 2.5686,
+      "step": 302200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.035222542427153e-05,
+      "loss": 2.5647,
+      "step": 302300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.034902337495998e-05,
+      "loss": 2.5552,
+      "step": 302400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.034582132564842e-05,
+      "loss": 2.5441,
+      "step": 302500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0342619276336856e-05,
+      "loss": 2.5753,
+      "step": 302600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.03394172270253e-05,
+      "loss": 2.5821,
+      "step": 302700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0336215177713736e-05,
+      "loss": 2.5517,
+      "step": 302800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.033301312840218e-05,
+      "loss": 2.5735,
+      "step": 302900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0329811079090615e-05,
+      "loss": 2.5704,
+      "step": 303000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.50827693939209,
+      "eval_runtime": 177.9887,
+      "eval_samples_per_second": 56.183,
+      "eval_steps_per_second": 3.511,
+      "step": 303000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.032660902977906e-05,
+      "loss": 2.5739,
+      "step": 303100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.03234069804675e-05,
+      "loss": 2.5599,
+      "step": 303200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.032020493115594e-05,
+      "loss": 2.5451,
+      "step": 303300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.031700288184439e-05,
+      "loss": 2.5776,
+      "step": 303400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.031380083253282e-05,
+      "loss": 2.5636,
+      "step": 303500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.031059878322127e-05,
+      "loss": 2.5535,
+      "step": 303600
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.03073967339097e-05,
+      "loss": 2.5527,
+      "step": 303700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.030419468459815e-05,
+      "loss": 2.5781,
+      "step": 303800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.030099263528658e-05,
+      "loss": 2.5719,
+      "step": 303900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0297790585975026e-05,
+      "loss": 2.5473,
+      "step": 304000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.5073838233947754,
+      "eval_runtime": 179.4958,
+      "eval_samples_per_second": 55.712,
+      "eval_steps_per_second": 3.482,
+      "step": 304000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0294588536663466e-05,
+      "loss": 2.548,
+      "step": 304100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0291386487351906e-05,
+      "loss": 2.5382,
+      "step": 304200
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.028818443804035e-05,
+      "loss": 2.5654,
+      "step": 304300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0284982388728785e-05,
+      "loss": 2.5579,
+      "step": 304400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.028178033941723e-05,
+      "loss": 2.5408,
+      "step": 304500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.0278578290105665e-05,
+      "loss": 2.564,
+      "step": 304600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.027537624079411e-05,
+      "loss": 2.5394,
+      "step": 304700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.027217419148255e-05,
+      "loss": 2.564,
+      "step": 304800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.026897214217099e-05,
+      "loss": 2.5578,
+      "step": 304900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.026577009285944e-05,
+      "loss": 2.5446,
+      "step": 305000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5069239139556885,
+      "eval_runtime": 176.0746,
+      "eval_samples_per_second": 56.794,
+      "eval_steps_per_second": 3.55,
+      "step": 305000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.026256804354787e-05,
+      "loss": 2.5529,
+      "step": 305100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.025936599423632e-05,
+      "loss": 2.5475,
+      "step": 305200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.025616394492475e-05,
+      "loss": 2.5522,
+      "step": 305300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0252961895613196e-05,
+      "loss": 2.5702,
+      "step": 305400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.024975984630163e-05,
+      "loss": 2.5735,
+      "step": 305500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0246557796990076e-05,
+      "loss": 2.5599,
+      "step": 305600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0243355747678515e-05,
+      "loss": 2.5524,
+      "step": 305700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0240153698366955e-05,
+      "loss": 2.5459,
+      "step": 305800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.02369516490554e-05,
+      "loss": 2.5709,
+      "step": 305900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0233749599743835e-05,
+      "loss": 2.5669,
+      "step": 306000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5048508644104004,
+      "eval_runtime": 177.2145,
+      "eval_samples_per_second": 56.429,
+      "eval_steps_per_second": 3.527,
+      "step": 306000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.023054755043228e-05,
+      "loss": 2.572,
+      "step": 306100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0227345501120714e-05,
+      "loss": 2.551,
+      "step": 306200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.022414345180916e-05,
+      "loss": 2.5737,
+      "step": 306300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.02209414024976e-05,
+      "loss": 2.5602,
+      "step": 306400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.021773935318604e-05,
+      "loss": 2.5354,
+      "step": 306500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0214537303874487e-05,
+      "loss": 2.5582,
+      "step": 306600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.021133525456292e-05,
+      "loss": 2.5414,
+      "step": 306700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0208133205251366e-05,
+      "loss": 2.5451,
+      "step": 306800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.02049311559398e-05,
+      "loss": 2.5318,
+      "step": 306900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0201729106628246e-05,
+      "loss": 2.5518,
+      "step": 307000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.507680892944336,
+      "eval_runtime": 177.91,
+      "eval_samples_per_second": 56.208,
+      "eval_steps_per_second": 3.513,
+      "step": 307000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0198527057316685e-05,
+      "loss": 2.5772,
+      "step": 307100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0195325008005125e-05,
+      "loss": 2.5669,
+      "step": 307200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0192122958693565e-05,
+      "loss": 2.5582,
+      "step": 307300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0188920909382004e-05,
+      "loss": 2.5566,
+      "step": 307400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.018571886007045e-05,
+      "loss": 2.5654,
+      "step": 307500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0182516810758884e-05,
+      "loss": 2.5824,
+      "step": 307600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.017931476144733e-05,
+      "loss": 2.5633,
+      "step": 307700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0176112712135763e-05,
+      "loss": 2.5698,
+      "step": 307800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.017291066282421e-05,
+      "loss": 2.542,
+      "step": 307900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.016970861351265e-05,
+      "loss": 2.5414,
+      "step": 308000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5082297325134277,
+      "eval_runtime": 174.5318,
+      "eval_samples_per_second": 57.296,
+      "eval_steps_per_second": 3.581,
+      "step": 308000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.016650656420109e-05,
+      "loss": 2.55,
+      "step": 308100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0163304514889536e-05,
+      "loss": 2.5611,
+      "step": 308200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.016010246557797e-05,
+      "loss": 2.5464,
+      "step": 308300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0156900416266415e-05,
+      "loss": 2.5544,
+      "step": 308400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.015369836695485e-05,
+      "loss": 2.5612,
+      "step": 308500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0150496317643295e-05,
+      "loss": 2.5487,
+      "step": 308600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0147294268331735e-05,
+      "loss": 2.5452,
+      "step": 308700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0144092219020174e-05,
+      "loss": 2.5652,
+      "step": 308800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0140890169708614e-05,
+      "loss": 2.5409,
+      "step": 308900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0137688120397054e-05,
+      "loss": 2.5588,
+      "step": 309000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5070741176605225,
+      "eval_runtime": 176.2748,
+      "eval_samples_per_second": 56.73,
+      "eval_steps_per_second": 3.546,
+      "step": 309000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.01344860710855e-05,
+      "loss": 2.553,
+      "step": 309100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.013128402177393e-05,
+      "loss": 2.5702,
+      "step": 309200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.012808197246238e-05,
+      "loss": 2.5501,
+      "step": 309300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.012487992315082e-05,
+      "loss": 2.5654,
+      "step": 309400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.012167787383926e-05,
+      "loss": 2.5569,
+      "step": 309500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.01184758245277e-05,
+      "loss": 2.5349,
+      "step": 309600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.011527377521614e-05,
+      "loss": 2.5506,
+      "step": 309700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0112071725904585e-05,
+      "loss": 2.5564,
+      "step": 309800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.010886967659302e-05,
+      "loss": 2.5588,
+      "step": 309900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0105667627281465e-05,
+      "loss": 2.5626,
+      "step": 310000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.506380796432495,
+      "eval_runtime": 181.6871,
+      "eval_samples_per_second": 55.04,
+      "eval_steps_per_second": 3.44,
+      "step": 310000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.01024655779699e-05,
+      "loss": 2.5418,
+      "step": 310100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0099263528658344e-05,
+      "loss": 2.5562,
+      "step": 310200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0096061479346784e-05,
+      "loss": 2.5452,
+      "step": 310300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0092859430035224e-05,
+      "loss": 2.5572,
+      "step": 310400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0089657380723663e-05,
+      "loss": 2.5605,
+      "step": 310500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.00864553314121e-05,
+      "loss": 2.5521,
+      "step": 310600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.008325328210055e-05,
+      "loss": 2.5529,
+      "step": 310700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.008005123278898e-05,
+      "loss": 2.5305,
+      "step": 310800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.007684918347743e-05,
+      "loss": 2.5354,
+      "step": 310900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.007364713416587e-05,
+      "loss": 2.5514,
+      "step": 311000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.505598545074463,
+      "eval_runtime": 174.1208,
+      "eval_samples_per_second": 57.431,
+      "eval_steps_per_second": 3.589,
+      "step": 311000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.007044508485431e-05,
+      "loss": 2.5501,
+      "step": 311100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.006724303554275e-05,
+      "loss": 2.5382,
+      "step": 311200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.006404098623119e-05,
+      "loss": 2.5473,
+      "step": 311300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0060838936919635e-05,
+      "loss": 2.5561,
+      "step": 311400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.005763688760807e-05,
+      "loss": 2.5504,
+      "step": 311500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0054434838296514e-05,
+      "loss": 2.5553,
+      "step": 311600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0051232788984954e-05,
+      "loss": 2.5465,
+      "step": 311700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0048030739673394e-05,
+      "loss": 2.5427,
+      "step": 311800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.004482869036183e-05,
+      "loss": 2.558,
+      "step": 311900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.004162664105027e-05,
+      "loss": 2.5363,
+      "step": 312000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.505476236343384,
+      "eval_runtime": 179.3041,
+      "eval_samples_per_second": 55.771,
+      "eval_steps_per_second": 3.486,
+      "step": 312000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.003842459173871e-05,
+      "loss": 2.5484,
+      "step": 312100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.003522254242715e-05,
+      "loss": 2.5404,
+      "step": 312200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.00320204931156e-05,
+      "loss": 2.5442,
+      "step": 312300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.002881844380404e-05,
+      "loss": 2.5239,
+      "step": 312400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.002561639449248e-05,
+      "loss": 2.5535,
+      "step": 312500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.002241434518092e-05,
+      "loss": 2.5565,
+      "step": 312600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.001921229586936e-05,
+      "loss": 2.5325,
+      "step": 312700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.00160102465578e-05,
+      "loss": 2.5471,
+      "step": 312800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.001280819724624e-05,
+      "loss": 2.5279,
+      "step": 312900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.000960614793468e-05,
+      "loss": 2.5293,
+      "step": 313000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.50457763671875,
+      "eval_runtime": 182.9787,
+      "eval_samples_per_second": 54.651,
+      "eval_steps_per_second": 3.416,
+      "step": 313000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.000640409862312e-05,
+      "loss": 2.5403,
+      "step": 313100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0003202049311564e-05,
+      "loss": 2.5394,
+      "step": 313200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4e-05,
+      "loss": 2.5552,
+      "step": 313300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.999679795068844e-05,
+      "loss": 2.5321,
+      "step": 313400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.999359590137688e-05,
+      "loss": 2.5459,
+      "step": 313500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.999039385206532e-05,
+      "loss": 2.5312,
+      "step": 313600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.998719180275376e-05,
+      "loss": 2.5396,
+      "step": 313700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.99839897534422e-05,
+      "loss": 2.5456,
+      "step": 313800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.998078770413065e-05,
+      "loss": 2.546,
+      "step": 313900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.997758565481909e-05,
+      "loss": 2.5397,
+      "step": 314000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5051333904266357,
+      "eval_runtime": 176.2002,
+      "eval_samples_per_second": 56.754,
+      "eval_steps_per_second": 3.547,
+      "step": 314000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.997438360550753e-05,
+      "loss": 2.5528,
+      "step": 314100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.997118155619597e-05,
+      "loss": 2.5509,
+      "step": 314200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.996797950688441e-05,
+      "loss": 2.5298,
+      "step": 314300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.996477745757285e-05,
+      "loss": 2.5405,
+      "step": 314400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.996157540826129e-05,
+      "loss": 2.5263,
+      "step": 314500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.995837335894973e-05,
+      "loss": 2.5482,
+      "step": 314600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.995517130963817e-05,
+      "loss": 2.5449,
+      "step": 314700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.995196926032661e-05,
+      "loss": 2.5438,
+      "step": 314800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.994876721101505e-05,
+      "loss": 2.5317,
+      "step": 314900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.994556516170349e-05,
+      "loss": 2.5511,
+      "step": 315000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5045018196105957,
+      "eval_runtime": 179.4239,
+      "eval_samples_per_second": 55.734,
+      "eval_steps_per_second": 3.483,
+      "step": 315000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.994236311239193e-05,
+      "loss": 2.5536,
+      "step": 315100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.993916106308037e-05,
+      "loss": 2.5616,
+      "step": 315200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.993595901376881e-05,
+      "loss": 2.5476,
+      "step": 315300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.993275696445725e-05,
+      "loss": 2.562,
+      "step": 315400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.99295549151457e-05,
+      "loss": 2.5341,
+      "step": 315500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.992635286583414e-05,
+      "loss": 2.5354,
+      "step": 315600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.992315081652258e-05,
+      "loss": 2.5496,
+      "step": 315700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.991994876721102e-05,
+      "loss": 2.5394,
+      "step": 315800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.991674671789946e-05,
+      "loss": 2.5321,
+      "step": 315900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9913544668587897e-05,
+      "loss": 2.5372,
+      "step": 316000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5045223236083984,
+      "eval_runtime": 176.428,
+      "eval_samples_per_second": 56.68,
+      "eval_steps_per_second": 3.543,
+      "step": 316000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9910342619276336e-05,
+      "loss": 2.5451,
+      "step": 316100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9907140569964776e-05,
+      "loss": 2.5537,
+      "step": 316200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.990393852065322e-05,
+      "loss": 2.558,
+      "step": 316300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.990073647134166e-05,
+      "loss": 2.5495,
+      "step": 316400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.98975344220301e-05,
+      "loss": 2.5321,
+      "step": 316500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.989433237271854e-05,
+      "loss": 2.535,
+      "step": 316600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.989113032340698e-05,
+      "loss": 2.5428,
+      "step": 316700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.988792827409542e-05,
+      "loss": 2.54,
+      "step": 316800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.988472622478386e-05,
+      "loss": 2.5508,
+      "step": 316900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.988152417547231e-05,
+      "loss": 2.5319,
+      "step": 317000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5040602684020996,
+      "eval_runtime": 174.4318,
+      "eval_samples_per_second": 57.329,
+      "eval_steps_per_second": 3.583,
+      "step": 317000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.987832212616075e-05,
+      "loss": 2.5404,
+      "step": 317100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.987512007684919e-05,
+      "loss": 2.5349,
+      "step": 317200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.987191802753763e-05,
+      "loss": 2.5611,
+      "step": 317300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9868715978226066e-05,
+      "loss": 2.5419,
+      "step": 317400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9865513928914506e-05,
+      "loss": 2.538,
+      "step": 317500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9862311879602946e-05,
+      "loss": 2.5602,
+      "step": 317600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9859109830291386e-05,
+      "loss": 2.5787,
+      "step": 317700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9855907780979825e-05,
+      "loss": 2.5444,
+      "step": 317800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.985270573166827e-05,
+      "loss": 2.559,
+      "step": 317900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.984950368235671e-05,
+      "loss": 2.5492,
+      "step": 318000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.504504919052124,
+      "eval_runtime": 177.5462,
+      "eval_samples_per_second": 56.323,
+      "eval_steps_per_second": 3.52,
+      "step": 318000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.984630163304515e-05,
+      "loss": 2.5417,
+      "step": 318100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.984309958373359e-05,
+      "loss": 2.5378,
+      "step": 318200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.983989753442203e-05,
+      "loss": 2.5486,
+      "step": 318300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.983669548511047e-05,
+      "loss": 2.5299,
+      "step": 318400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.983349343579891e-05,
+      "loss": 2.5635,
+      "step": 318500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.983029138648736e-05,
+      "loss": 2.559,
+      "step": 318600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9827089337175797e-05,
+      "loss": 2.5321,
+      "step": 318700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9823887287864236e-05,
+      "loss": 2.5433,
+      "step": 318800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9820685238552676e-05,
+      "loss": 2.5332,
+      "step": 318900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9817483189241116e-05,
+      "loss": 2.5496,
+      "step": 319000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5038700103759766,
+      "eval_runtime": 180.8354,
+      "eval_samples_per_second": 55.299,
+      "eval_steps_per_second": 3.456,
+      "step": 319000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9814281139929556e-05,
+      "loss": 2.5304,
+      "step": 319100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9811079090617995e-05,
+      "loss": 2.5197,
+      "step": 319200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.980787704130644e-05,
+      "loss": 2.5367,
+      "step": 319300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9804674991994875e-05,
+      "loss": 2.5455,
+      "step": 319400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.980147294268332e-05,
+      "loss": 2.5495,
+      "step": 319500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.979827089337176e-05,
+      "loss": 2.544,
+      "step": 319600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.97950688440602e-05,
+      "loss": 2.5515,
+      "step": 319700
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.979186679474864e-05,
+      "loss": 2.5458,
+      "step": 319800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.978866474543708e-05,
+      "loss": 2.5365,
+      "step": 319900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.978546269612553e-05,
+      "loss": 2.5437,
+      "step": 320000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.5052034854888916,
+      "eval_runtime": 177.9273,
+      "eval_samples_per_second": 56.203,
+      "eval_steps_per_second": 3.513,
+      "step": 320000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.978226064681396e-05,
+      "loss": 2.5319,
+      "step": 320100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9779058597502406e-05,
+      "loss": 2.5411,
+      "step": 320200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 3.9775856548190846e-05,
+      "loss": 2.5497,
+      "step": 320300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9772654498879286e-05,
+      "loss": 2.5301,
+      "step": 320400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9769452449567725e-05,
+      "loss": 2.5616,
+      "step": 320500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9766250400256165e-05,
+      "loss": 2.5533,
+      "step": 320600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9763048350944605e-05,
+      "loss": 2.5563,
+      "step": 320700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9759846301633045e-05,
+      "loss": 2.535,
+      "step": 320800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.975664425232149e-05,
+      "loss": 2.5515,
+      "step": 320900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9753442203009924e-05,
+      "loss": 2.5478,
+      "step": 321000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.502642869949341,
+      "eval_runtime": 176.3063,
+      "eval_samples_per_second": 56.719,
+      "eval_steps_per_second": 3.545,
+      "step": 321000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.975024015369837e-05,
+      "loss": 2.5499,
+      "step": 321100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.974703810438681e-05,
+      "loss": 2.5389,
+      "step": 321200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.974383605507525e-05,
+      "loss": 2.5296,
+      "step": 321300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.974063400576369e-05,
+      "loss": 2.5517,
+      "step": 321400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.973743195645213e-05,
+      "loss": 2.5395,
+      "step": 321500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9734229907140576e-05,
+      "loss": 2.5425,
+      "step": 321600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.973102785782901e-05,
+      "loss": 2.5416,
+      "step": 321700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9727825808517456e-05,
+      "loss": 2.5388,
+      "step": 321800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9724623759205895e-05,
+      "loss": 2.5441,
+      "step": 321900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9721421709894335e-05,
+      "loss": 2.5368,
+      "step": 322000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.505239725112915,
+      "eval_runtime": 176.1017,
+      "eval_samples_per_second": 56.785,
+      "eval_steps_per_second": 3.549,
+      "step": 322000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9718219660582775e-05,
+      "loss": 2.5253,
+      "step": 322100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9715017611271215e-05,
+      "loss": 2.558,
+      "step": 322200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.971181556195966e-05,
+      "loss": 2.5313,
+      "step": 322300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9708613512648094e-05,
+      "loss": 2.5337,
+      "step": 322400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.970541146333654e-05,
+      "loss": 2.5383,
+      "step": 322500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9702209414024974e-05,
+      "loss": 2.5247,
+      "step": 322600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.969900736471342e-05,
+      "loss": 2.5411,
+      "step": 322700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.969580531540186e-05,
+      "loss": 2.558,
+      "step": 322800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.96926032660903e-05,
+      "loss": 2.5671,
+      "step": 322900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.968940121677874e-05,
+      "loss": 2.5209,
+      "step": 323000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.504387617111206,
+      "eval_runtime": 176.9359,
+      "eval_samples_per_second": 56.518,
+      "eval_steps_per_second": 3.532,
+      "step": 323000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.968619916746718e-05,
+      "loss": 2.5459,
+      "step": 323100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9682997118155625e-05,
+      "loss": 2.5554,
+      "step": 323200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.967979506884406e-05,
+      "loss": 2.5532,
+      "step": 323300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9676593019532505e-05,
+      "loss": 2.5303,
+      "step": 323400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9673390970220945e-05,
+      "loss": 2.5224,
+      "step": 323500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9670188920909384e-05,
+      "loss": 2.5402,
+      "step": 323600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9666986871597824e-05,
+      "loss": 2.5362,
+      "step": 323700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9663784822286264e-05,
+      "loss": 2.5348,
+      "step": 323800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.966058277297471e-05,
+      "loss": 2.5452,
+      "step": 323900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9657380723663143e-05,
+      "loss": 2.5453,
+      "step": 324000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.503420829772949,
+      "eval_runtime": 175.8923,
+      "eval_samples_per_second": 56.853,
+      "eval_steps_per_second": 3.553,
+      "step": 324000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.965417867435159e-05,
+      "loss": 2.5733,
+      "step": 324100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.965097662504002e-05,
+      "loss": 2.5448,
+      "step": 324200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.964777457572847e-05,
+      "loss": 2.5223,
+      "step": 324300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.964457252641691e-05,
+      "loss": 2.5542,
+      "step": 324400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.964137047710535e-05,
+      "loss": 2.5298,
+      "step": 324500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9638168427793795e-05,
+      "loss": 2.5369,
+      "step": 324600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.963496637848223e-05,
+      "loss": 2.5425,
+      "step": 324700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9631764329170675e-05,
+      "loss": 2.547,
+      "step": 324800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.962856227985911e-05,
+      "loss": 2.5343,
+      "step": 324900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9625360230547554e-05,
+      "loss": 2.5714,
+      "step": 325000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.503641366958618,
+      "eval_runtime": 174.1884,
+      "eval_samples_per_second": 57.409,
+      "eval_steps_per_second": 3.588,
+      "step": 325000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9622158181235994e-05,
+      "loss": 2.5426,
+      "step": 325100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9618956131924434e-05,
+      "loss": 2.537,
+      "step": 325200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9615754082612874e-05,
+      "loss": 2.5445,
+      "step": 325300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.961255203330131e-05,
+      "loss": 2.5471,
+      "step": 325400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.960934998398976e-05,
+      "loss": 2.546,
+      "step": 325500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.960614793467819e-05,
+      "loss": 2.537,
+      "step": 325600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.960294588536664e-05,
+      "loss": 2.5471,
+      "step": 325700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.959974383605507e-05,
+      "loss": 2.5305,
+      "step": 325800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.959654178674352e-05,
+      "loss": 2.5455,
+      "step": 325900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.959333973743196e-05,
+      "loss": 2.5335,
+      "step": 326000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.5042974948883057,
+      "eval_runtime": 176.1528,
+      "eval_samples_per_second": 56.769,
+      "eval_steps_per_second": 3.548,
+      "step": 326000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.95901376881204e-05,
+      "loss": 2.5461,
+      "step": 326100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9586935638808845e-05,
+      "loss": 2.5605,
+      "step": 326200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.958373358949728e-05,
+      "loss": 2.5226,
+      "step": 326300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9580531540185724e-05,
+      "loss": 2.5242,
+      "step": 326400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.957732949087416e-05,
+      "loss": 2.5546,
+      "step": 326500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9574127441562604e-05,
+      "loss": 2.544,
+      "step": 326600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9570925392251043e-05,
+      "loss": 2.5166,
+      "step": 326700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.956772334293948e-05,
+      "loss": 2.5471,
+      "step": 326800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.956452129362793e-05,
+      "loss": 2.5462,
+      "step": 326900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.956131924431636e-05,
+      "loss": 2.5419,
+      "step": 327000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.502007246017456,
+      "eval_runtime": 181.6416,
+      "eval_samples_per_second": 55.053,
+      "eval_steps_per_second": 3.441,
+      "step": 327000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.955811719500481e-05,
+      "loss": 2.5315,
+      "step": 327100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.955491514569324e-05,
+      "loss": 2.5304,
+      "step": 327200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.955171309638169e-05,
+      "loss": 2.5687,
+      "step": 327300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.954851104707012e-05,
+      "loss": 2.5477,
+      "step": 327400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.954530899775857e-05,
+      "loss": 2.5408,
+      "step": 327500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.954210694844701e-05,
+      "loss": 2.5411,
+      "step": 327600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.953890489913545e-05,
+      "loss": 2.5467,
+      "step": 327700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9535702849823894e-05,
+      "loss": 2.548,
+      "step": 327800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.953250080051233e-05,
+      "loss": 2.5375,
+      "step": 327900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9529298751200774e-05,
+      "loss": 2.554,
+      "step": 328000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.504699468612671,
+      "eval_runtime": 173.722,
+      "eval_samples_per_second": 57.563,
+      "eval_steps_per_second": 3.598,
+      "step": 328000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9526096701889207e-05,
+      "loss": 2.5319,
+      "step": 328100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.952289465257765e-05,
+      "loss": 2.5416,
+      "step": 328200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.951969260326609e-05,
+      "loss": 2.5422,
+      "step": 328300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.951649055395453e-05,
+      "loss": 2.5598,
+      "step": 328400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.951328850464298e-05,
+      "loss": 2.5542,
+      "step": 328500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.951008645533141e-05,
+      "loss": 2.5342,
+      "step": 328600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.950688440601986e-05,
+      "loss": 2.5362,
+      "step": 328700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.950368235670829e-05,
+      "loss": 2.5397,
+      "step": 328800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.950048030739674e-05,
+      "loss": 2.5425,
+      "step": 328900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.949727825808517e-05,
+      "loss": 2.5315,
+      "step": 329000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.5056402683258057,
+      "eval_runtime": 177.4283,
+      "eval_samples_per_second": 56.361,
+      "eval_steps_per_second": 3.523,
+      "step": 329000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.949407620877362e-05,
+      "loss": 2.5413,
+      "step": 329100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.949087415946206e-05,
+      "loss": 2.5423,
+      "step": 329200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.94876721101505e-05,
+      "loss": 2.5424,
+      "step": 329300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9484470060838944e-05,
+      "loss": 2.5236,
+      "step": 329400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9481268011527376e-05,
+      "loss": 2.5192,
+      "step": 329500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.947806596221582e-05,
+      "loss": 2.5476,
+      "step": 329600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9474863912904256e-05,
+      "loss": 2.5511,
+      "step": 329700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.94716618635927e-05,
+      "loss": 2.5232,
+      "step": 329800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.946845981428114e-05,
+      "loss": 2.5542,
+      "step": 329900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.946525776496958e-05,
+      "loss": 2.535,
+      "step": 330000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.508066177368164,
+      "eval_runtime": 182.1861,
+      "eval_samples_per_second": 54.889,
+      "eval_steps_per_second": 3.431,
+      "step": 330000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.946205571565802e-05,
+      "loss": 2.5529,
+      "step": 330100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.945885366634646e-05,
+      "loss": 2.5257,
+      "step": 330200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.945565161703491e-05,
+      "loss": 2.5365,
+      "step": 330300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.945244956772334e-05,
+      "loss": 2.5432,
+      "step": 330400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.944924751841179e-05,
+      "loss": 2.5478,
+      "step": 330500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.944604546910022e-05,
+      "loss": 2.5398,
+      "step": 330600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.944284341978867e-05,
+      "loss": 2.5203,
+      "step": 330700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.943964137047711e-05,
+      "loss": 2.5287,
+      "step": 330800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9436439321165546e-05,
+      "loss": 2.5378,
+      "step": 330900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.943323727185399e-05,
+      "loss": 2.5431,
+      "step": 331000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.5063817501068115,
+      "eval_runtime": 177.3593,
+      "eval_samples_per_second": 56.383,
+      "eval_steps_per_second": 3.524,
+      "step": 331000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9430035222542426e-05,
+      "loss": 2.5348,
+      "step": 331100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.942683317323087e-05,
+      "loss": 2.5277,
+      "step": 331200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9423631123919305e-05,
+      "loss": 2.5345,
+      "step": 331300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.942042907460775e-05,
+      "loss": 2.5292,
+      "step": 331400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.941722702529619e-05,
+      "loss": 2.5705,
+      "step": 331500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.941402497598463e-05,
+      "loss": 2.5488,
+      "step": 331600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.941082292667307e-05,
+      "loss": 2.5694,
+      "step": 331700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.940762087736151e-05,
+      "loss": 2.5491,
+      "step": 331800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.940441882804996e-05,
+      "loss": 2.5307,
+      "step": 331900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.940121677873839e-05,
+      "loss": 2.5362,
+      "step": 332000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.5041511058807373,
+      "eval_runtime": 173.3732,
+      "eval_samples_per_second": 57.679,
+      "eval_steps_per_second": 3.605,
+      "step": 332000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.939801472942684e-05,
+      "loss": 2.5468,
+      "step": 332100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9394812680115277e-05,
+      "loss": 2.5429,
+      "step": 332200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9391610630803716e-05,
+      "loss": 2.5514,
+      "step": 332300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9388408581492156e-05,
+      "loss": 2.5171,
+      "step": 332400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9385206532180596e-05,
+      "loss": 2.5341,
+      "step": 332500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.938200448286904e-05,
+      "loss": 2.5426,
+      "step": 332600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9378802433557475e-05,
+      "loss": 2.5406,
+      "step": 332700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.937560038424592e-05,
+      "loss": 2.5297,
+      "step": 332800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9372398334934355e-05,
+      "loss": 2.5386,
+      "step": 332900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.93691962856228e-05,
+      "loss": 2.5313,
+      "step": 333000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.5052037239074707,
+      "eval_runtime": 178.1571,
+      "eval_samples_per_second": 56.13,
+      "eval_steps_per_second": 3.508,
+      "step": 333000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.936599423631124e-05,
+      "loss": 2.5314,
+      "step": 333100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.936279218699968e-05,
+      "loss": 2.5428,
+      "step": 333200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.935959013768812e-05,
+      "loss": 2.539,
+      "step": 333300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.935638808837656e-05,
+      "loss": 2.5373,
+      "step": 333400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.935318603906501e-05,
+      "loss": 2.5497,
+      "step": 333500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.934998398975344e-05,
+      "loss": 2.5336,
+      "step": 333600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9346781940441886e-05,
+      "loss": 2.5407,
+      "step": 333700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9343579891130326e-05,
+      "loss": 2.5352,
+      "step": 333800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9340377841818766e-05,
+      "loss": 2.5255,
+      "step": 333900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9337175792507205e-05,
+      "loss": 2.5323,
+      "step": 334000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.5034334659576416,
+      "eval_runtime": 177.3628,
+      "eval_samples_per_second": 56.382,
+      "eval_steps_per_second": 3.524,
+      "step": 334000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9333973743195645e-05,
+      "loss": 2.5344,
+      "step": 334100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.933077169388409e-05,
+      "loss": 2.5478,
+      "step": 334200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9327569644572525e-05,
+      "loss": 2.5514,
+      "step": 334300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.932436759526097e-05,
+      "loss": 2.5434,
+      "step": 334400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.932116554594941e-05,
+      "loss": 2.5329,
+      "step": 334500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.931796349663785e-05,
+      "loss": 2.5589,
+      "step": 334600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.931476144732629e-05,
+      "loss": 2.5424,
+      "step": 334700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.931155939801473e-05,
+      "loss": 2.5403,
+      "step": 334800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.930835734870317e-05,
+      "loss": 2.5382,
+      "step": 334900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.930515529939161e-05,
+      "loss": 2.5531,
+      "step": 335000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.502439260482788,
+      "eval_runtime": 177.2654,
+      "eval_samples_per_second": 56.413,
+      "eval_steps_per_second": 3.526,
+      "step": 335000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9301953250080056e-05,
+      "loss": 2.5458,
+      "step": 335100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.929875120076849e-05,
+      "loss": 2.5491,
+      "step": 335200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9295549151456936e-05,
+      "loss": 2.5313,
+      "step": 335300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9292347102145375e-05,
+      "loss": 2.5275,
+      "step": 335400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9289145052833815e-05,
+      "loss": 2.5285,
+      "step": 335500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9285943003522255e-05,
+      "loss": 2.527,
+      "step": 335600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9282740954210694e-05,
+      "loss": 2.519,
+      "step": 335700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.927953890489914e-05,
+      "loss": 2.5079,
+      "step": 335800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 3.9276336855587574e-05,
+      "loss": 2.5293,
+      "step": 335900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.927313480627602e-05,
+      "loss": 2.534,
+      "step": 336000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.505464553833008,
+      "eval_runtime": 179.7562,
+      "eval_samples_per_second": 55.631,
+      "eval_steps_per_second": 3.477,
+      "step": 336000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.926993275696446e-05,
+      "loss": 2.5273,
+      "step": 336100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.92667307076529e-05,
+      "loss": 2.5203,
+      "step": 336200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.926352865834134e-05,
+      "loss": 2.5347,
+      "step": 336300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.926032660902978e-05,
+      "loss": 2.5287,
+      "step": 336400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.925712455971822e-05,
+      "loss": 2.5166,
+      "step": 336500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.925392251040666e-05,
+      "loss": 2.5404,
+      "step": 336600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9250720461095105e-05,
+      "loss": 2.5171,
+      "step": 336700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9247518411783545e-05,
+      "loss": 2.5414,
+      "step": 336800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9244316362471985e-05,
+      "loss": 2.5476,
+      "step": 336900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9241114313160425e-05,
+      "loss": 2.5311,
+      "step": 337000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.503237009048462,
+      "eval_runtime": 177.8754,
+      "eval_samples_per_second": 56.219,
+      "eval_steps_per_second": 3.514,
+      "step": 337000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9237912263848864e-05,
+      "loss": 2.5438,
+      "step": 337100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9234710214537304e-05,
+      "loss": 2.5231,
+      "step": 337200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9231508165225744e-05,
+      "loss": 2.5314,
+      "step": 337300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.922830611591419e-05,
+      "loss": 2.5164,
+      "step": 337400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.922510406660262e-05,
+      "loss": 2.5187,
+      "step": 337500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.922190201729107e-05,
+      "loss": 2.5424,
+      "step": 337600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.921869996797951e-05,
+      "loss": 2.5342,
+      "step": 337700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.921549791866795e-05,
+      "loss": 2.5107,
+      "step": 337800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.921229586935639e-05,
+      "loss": 2.5053,
+      "step": 337900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.920909382004483e-05,
+      "loss": 2.537,
+      "step": 338000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.502249002456665,
+      "eval_runtime": 177.7412,
+      "eval_samples_per_second": 56.262,
+      "eval_steps_per_second": 3.516,
+      "step": 338000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.920589177073327e-05,
+      "loss": 2.5133,
+      "step": 338100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.920268972142171e-05,
+      "loss": 2.5349,
+      "step": 338200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9199487672110155e-05,
+      "loss": 2.5304,
+      "step": 338300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9196285622798595e-05,
+      "loss": 2.5281,
+      "step": 338400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9193083573487034e-05,
+      "loss": 2.5396,
+      "step": 338500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9189881524175474e-05,
+      "loss": 2.5307,
+      "step": 338600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9186679474863914e-05,
+      "loss": 2.5417,
+      "step": 338700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9183477425552353e-05,
+      "loss": 2.5217,
+      "step": 338800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.918027537624079e-05,
+      "loss": 2.5489,
+      "step": 338900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.917707332692924e-05,
+      "loss": 2.5204,
+      "step": 339000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.5009543895721436,
+      "eval_runtime": 177.5956,
+      "eval_samples_per_second": 56.308,
+      "eval_steps_per_second": 3.519,
+      "step": 339000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.917387127761768e-05,
+      "loss": 2.5157,
+      "step": 339100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.917066922830612e-05,
+      "loss": 2.5283,
+      "step": 339200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.916746717899456e-05,
+      "loss": 2.4974,
+      "step": 339300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9164265129683e-05,
+      "loss": 2.5386,
+      "step": 339400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.916106308037144e-05,
+      "loss": 2.5457,
+      "step": 339500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.915786103105988e-05,
+      "loss": 2.5345,
+      "step": 339600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.915465898174832e-05,
+      "loss": 2.519,
+      "step": 339700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9151456932436764e-05,
+      "loss": 2.5253,
+      "step": 339800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9148254883125204e-05,
+      "loss": 2.5485,
+      "step": 339900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9145052833813644e-05,
+      "loss": 2.543,
+      "step": 340000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.5014827251434326,
+      "eval_runtime": 173.8552,
+      "eval_samples_per_second": 57.519,
+      "eval_steps_per_second": 3.595,
+      "step": 340000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9141850784502084e-05,
+      "loss": 2.5324,
+      "step": 340100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.913864873519052e-05,
+      "loss": 2.5207,
+      "step": 340200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.913544668587896e-05,
+      "loss": 2.5121,
+      "step": 340300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.91322446365674e-05,
+      "loss": 2.5151,
+      "step": 340400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.912904258725584e-05,
+      "loss": 2.5572,
+      "step": 340500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.912584053794429e-05,
+      "loss": 2.5336,
+      "step": 340600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.912263848863273e-05,
+      "loss": 2.5318,
+      "step": 340700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.911943643932117e-05,
+      "loss": 2.5118,
+      "step": 340800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.911623439000961e-05,
+      "loss": 2.5269,
+      "step": 340900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.911303234069805e-05,
+      "loss": 2.5227,
+      "step": 341000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.5016543865203857,
+      "eval_runtime": 180.0431,
+      "eval_samples_per_second": 55.542,
+      "eval_steps_per_second": 3.471,
+      "step": 341000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.910983029138649e-05,
+      "loss": 2.5317,
+      "step": 341100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.910662824207493e-05,
+      "loss": 2.5284,
+      "step": 341200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.910342619276337e-05,
+      "loss": 2.5298,
+      "step": 341300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9100224143451814e-05,
+      "loss": 2.5419,
+      "step": 341400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9097022094140254e-05,
+      "loss": 2.5179,
+      "step": 341500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.909382004482869e-05,
+      "loss": 2.509,
+      "step": 341600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.909061799551713e-05,
+      "loss": 2.5208,
+      "step": 341700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.908741594620557e-05,
+      "loss": 2.5141,
+      "step": 341800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.908421389689401e-05,
+      "loss": 2.5313,
+      "step": 341900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.908101184758245e-05,
+      "loss": 2.5236,
+      "step": 342000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.502887010574341,
+      "eval_runtime": 177.9793,
+      "eval_samples_per_second": 56.186,
+      "eval_steps_per_second": 3.512,
+      "step": 342000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.90778097982709e-05,
+      "loss": 2.5436,
+      "step": 342100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.907460774895934e-05,
+      "loss": 2.5374,
+      "step": 342200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.907140569964778e-05,
+      "loss": 2.5199,
+      "step": 342300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.906820365033622e-05,
+      "loss": 2.5138,
+      "step": 342400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.906500160102466e-05,
+      "loss": 2.5207,
+      "step": 342500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.90617995517131e-05,
+      "loss": 2.544,
+      "step": 342600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.905859750240154e-05,
+      "loss": 2.5205,
+      "step": 342700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.905539545308998e-05,
+      "loss": 2.5191,
+      "step": 342800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.905219340377842e-05,
+      "loss": 2.501,
+      "step": 342900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.904899135446686e-05,
+      "loss": 2.5272,
+      "step": 343000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.5032293796539307,
+      "eval_runtime": 176.8306,
+      "eval_samples_per_second": 56.551,
+      "eval_steps_per_second": 3.534,
+      "step": 343000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.90457893051553e-05,
+      "loss": 2.5399,
+      "step": 343100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.904258725584374e-05,
+      "loss": 2.5317,
+      "step": 343200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.903938520653218e-05,
+      "loss": 2.5491,
+      "step": 343300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.903618315722062e-05,
+      "loss": 2.5377,
+      "step": 343400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.903298110790906e-05,
+      "loss": 2.5159,
+      "step": 343500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.90297790585975e-05,
+      "loss": 2.5141,
+      "step": 343600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.902657700928595e-05,
+      "loss": 2.5237,
+      "step": 343700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.902337495997439e-05,
+      "loss": 2.5061,
+      "step": 343800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.902017291066283e-05,
+      "loss": 2.5459,
+      "step": 343900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.901697086135127e-05,
+      "loss": 2.5172,
+      "step": 344000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.5020711421966553,
+      "eval_runtime": 174.1296,
+      "eval_samples_per_second": 57.429,
+      "eval_steps_per_second": 3.589,
+      "step": 344000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.901376881203971e-05,
+      "loss": 2.5239,
+      "step": 344100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.901056676272815e-05,
+      "loss": 2.5312,
+      "step": 344200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9007364713416587e-05,
+      "loss": 2.5279,
+      "step": 344300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.900416266410503e-05,
+      "loss": 2.5287,
+      "step": 344400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9000960614793466e-05,
+      "loss": 2.5267,
+      "step": 344500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.899775856548191e-05,
+      "loss": 2.5315,
+      "step": 344600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.899455651617035e-05,
+      "loss": 2.5198,
+      "step": 344700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.899135446685879e-05,
+      "loss": 2.5358,
+      "step": 344800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.898815241754723e-05,
+      "loss": 2.5314,
+      "step": 344900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.898495036823567e-05,
+      "loss": 2.5102,
+      "step": 345000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.500669240951538,
+      "eval_runtime": 174.1005,
+      "eval_samples_per_second": 57.438,
+      "eval_steps_per_second": 3.59,
+      "step": 345000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.898174831892411e-05,
+      "loss": 2.5264,
+      "step": 345100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.897854626961255e-05,
+      "loss": 2.4822,
+      "step": 345200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8975344220301e-05,
+      "loss": 2.5123,
+      "step": 345300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.897214217098944e-05,
+      "loss": 2.5433,
+      "step": 345400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.896894012167788e-05,
+      "loss": 2.5241,
+      "step": 345500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.896573807236632e-05,
+      "loss": 2.5093,
+      "step": 345600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8962536023054756e-05,
+      "loss": 2.5274,
+      "step": 345700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8959333973743196e-05,
+      "loss": 2.525,
+      "step": 345800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8956131924431636e-05,
+      "loss": 2.5402,
+      "step": 345900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.895292987512008e-05,
+      "loss": 2.5156,
+      "step": 346000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.5017688274383545,
+      "eval_runtime": 182.3704,
+      "eval_samples_per_second": 54.833,
+      "eval_steps_per_second": 3.427,
+      "step": 346000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8949727825808515e-05,
+      "loss": 2.5155,
+      "step": 346100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.894652577649696e-05,
+      "loss": 2.5315,
+      "step": 346200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.89433237271854e-05,
+      "loss": 2.5151,
+      "step": 346300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.894012167787384e-05,
+      "loss": 2.5185,
+      "step": 346400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.893691962856228e-05,
+      "loss": 2.5272,
+      "step": 346500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.893371757925072e-05,
+      "loss": 2.5344,
+      "step": 346600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.893051552993917e-05,
+      "loss": 2.5144,
+      "step": 346700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.89273134806276e-05,
+      "loss": 2.5232,
+      "step": 346800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.892411143131605e-05,
+      "loss": 2.5326,
+      "step": 346900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8920909382004487e-05,
+      "loss": 2.5229,
+      "step": 347000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.502462387084961,
+      "eval_runtime": 177.3184,
+      "eval_samples_per_second": 56.396,
+      "eval_steps_per_second": 3.525,
+      "step": 347000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8917707332692926e-05,
+      "loss": 2.5204,
+      "step": 347100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8914505283381366e-05,
+      "loss": 2.5095,
+      "step": 347200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8911303234069806e-05,
+      "loss": 2.5313,
+      "step": 347300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8908101184758246e-05,
+      "loss": 2.5229,
+      "step": 347400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8904899135446685e-05,
+      "loss": 2.503,
+      "step": 347500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.890169708613513e-05,
+      "loss": 2.5152,
+      "step": 347600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8898495036823565e-05,
+      "loss": 2.5295,
+      "step": 347700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.889529298751201e-05,
+      "loss": 2.5384,
+      "step": 347800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.889209093820045e-05,
+      "loss": 2.5224,
+      "step": 347900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 2.5292,
+      "step": 348000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.500412940979004,
+      "eval_runtime": 176.5813,
+      "eval_samples_per_second": 56.631,
+      "eval_steps_per_second": 3.539,
+      "step": 348000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.888568683957733e-05,
+      "loss": 2.5116,
+      "step": 348100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.888248479026577e-05,
+      "loss": 2.4974,
+      "step": 348200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.887928274095422e-05,
+      "loss": 2.5153,
+      "step": 348300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.887608069164265e-05,
+      "loss": 2.5223,
+      "step": 348400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8872878642331096e-05,
+      "loss": 2.5173,
+      "step": 348500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8869676593019536e-05,
+      "loss": 2.5406,
+      "step": 348600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8866474543707976e-05,
+      "loss": 2.5215,
+      "step": 348700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8863272494396415e-05,
+      "loss": 2.5127,
+      "step": 348800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8860070445084855e-05,
+      "loss": 2.5301,
+      "step": 348900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.88568683957733e-05,
+      "loss": 2.5334,
+      "step": 349000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.4996371269226074,
+      "eval_runtime": 173.9569,
+      "eval_samples_per_second": 57.485,
+      "eval_steps_per_second": 3.593,
+      "step": 349000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8853666346461735e-05,
+      "loss": 2.5127,
+      "step": 349100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.885046429715018e-05,
+      "loss": 2.4993,
+      "step": 349200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8847262247838614e-05,
+      "loss": 2.5131,
+      "step": 349300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.884406019852706e-05,
+      "loss": 2.5099,
+      "step": 349400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.88408581492155e-05,
+      "loss": 2.5165,
+      "step": 349500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.883765609990394e-05,
+      "loss": 2.5212,
+      "step": 349600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.883445405059239e-05,
+      "loss": 2.5353,
+      "step": 349700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.883125200128082e-05,
+      "loss": 2.5002,
+      "step": 349800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8828049951969266e-05,
+      "loss": 2.5112,
+      "step": 349900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.88248479026577e-05,
+      "loss": 2.5168,
+      "step": 350000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.4998345375061035,
+      "eval_runtime": 175.5683,
+      "eval_samples_per_second": 56.958,
+      "eval_steps_per_second": 3.56,
+      "step": 350000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8821645853346146e-05,
+      "loss": 2.5114,
+      "step": 350100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8818443804034585e-05,
+      "loss": 2.512,
+      "step": 350200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8815241754723025e-05,
+      "loss": 2.5388,
+      "step": 350300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8812039705411465e-05,
+      "loss": 2.5288,
+      "step": 350400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8808837656099905e-05,
+      "loss": 2.5107,
+      "step": 350500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.880563560678835e-05,
+      "loss": 2.5228,
+      "step": 350600
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8802433557476784e-05,
+      "loss": 2.5256,
+      "step": 350700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.879923150816523e-05,
+      "loss": 2.5209,
+      "step": 350800
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8796029458853664e-05,
+      "loss": 2.5366,
+      "step": 350900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.879282740954211e-05,
+      "loss": 2.5308,
+      "step": 351000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.5002245903015137,
+      "eval_runtime": 179.2279,
+      "eval_samples_per_second": 55.795,
+      "eval_steps_per_second": 3.487,
+      "step": 351000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.878962536023055e-05,
+      "loss": 2.5232,
+      "step": 351100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.878642331091899e-05,
+      "loss": 2.5057,
+      "step": 351200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8783221261607436e-05,
+      "loss": 2.534,
+      "step": 351300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.878001921229587e-05,
+      "loss": 2.5405,
+      "step": 351400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.8776817162984315e-05,
+      "loss": 2.5292,
+      "step": 351500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.877361511367275e-05,
+      "loss": 2.5251,
+      "step": 351600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8770413064361195e-05,
+      "loss": 2.5063,
+      "step": 351700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8767211015049635e-05,
+      "loss": 2.5231,
+      "step": 351800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8764008965738074e-05,
+      "loss": 2.505,
+      "step": 351900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8760806916426514e-05,
+      "loss": 2.5168,
+      "step": 352000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.5032877922058105,
+      "eval_runtime": 176.5268,
+      "eval_samples_per_second": 56.649,
+      "eval_steps_per_second": 3.541,
+      "step": 352000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8757604867114954e-05,
+      "loss": 2.4991,
+      "step": 352100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.87544028178034e-05,
+      "loss": 2.5305,
+      "step": 352200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8751200768491833e-05,
+      "loss": 2.5349,
+      "step": 352300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.874799871918028e-05,
+      "loss": 2.5165,
+      "step": 352400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.874479666986871e-05,
+      "loss": 2.498,
+      "step": 352500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.874159462055716e-05,
+      "loss": 2.5275,
+      "step": 352600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.87383925712456e-05,
+      "loss": 2.5178,
+      "step": 352700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.873519052193404e-05,
+      "loss": 2.5061,
+      "step": 352800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8731988472622485e-05,
+      "loss": 2.5156,
+      "step": 352900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.872878642331092e-05,
+      "loss": 2.5166,
+      "step": 353000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.500791072845459,
+      "eval_runtime": 179.6206,
+      "eval_samples_per_second": 55.673,
+      "eval_steps_per_second": 3.48,
+      "step": 353000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8725584373999365e-05,
+      "loss": 2.5127,
+      "step": 353100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.87223823246878e-05,
+      "loss": 2.5216,
+      "step": 353200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8719180275376244e-05,
+      "loss": 2.539,
+      "step": 353300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8715978226064684e-05,
+      "loss": 2.5005,
+      "step": 353400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8712776176753124e-05,
+      "loss": 2.5216,
+      "step": 353500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8709574127441564e-05,
+      "loss": 2.52,
+      "step": 353600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.870637207813e-05,
+      "loss": 2.5194,
+      "step": 353700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.870317002881845e-05,
+      "loss": 2.5032,
+      "step": 353800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.869996797950688e-05,
+      "loss": 2.5008,
+      "step": 353900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.869676593019533e-05,
+      "loss": 2.5085,
+      "step": 354000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.5006790161132812,
+      "eval_runtime": 175.044,
+      "eval_samples_per_second": 57.128,
+      "eval_steps_per_second": 3.571,
+      "step": 354000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.869356388088376e-05,
+      "loss": 2.5099,
+      "step": 354100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.869036183157221e-05,
+      "loss": 2.509,
+      "step": 354200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.868715978226065e-05,
+      "loss": 2.5359,
+      "step": 354300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.868395773294909e-05,
+      "loss": 2.5108,
+      "step": 354400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8680755683637535e-05,
+      "loss": 2.5438,
+      "step": 354500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.867755363432597e-05,
+      "loss": 2.5437,
+      "step": 354600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8674351585014414e-05,
+      "loss": 2.5349,
+      "step": 354700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.867114953570285e-05,
+      "loss": 2.5286,
+      "step": 354800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8667947486391294e-05,
+      "loss": 2.5114,
+      "step": 354900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8664745437079733e-05,
+      "loss": 2.5082,
+      "step": 355000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.5020599365234375,
+      "eval_runtime": 176.4383,
+      "eval_samples_per_second": 56.677,
+      "eval_steps_per_second": 3.542,
+      "step": 355000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.866154338776817e-05,
+      "loss": 2.5269,
+      "step": 355100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.865834133845661e-05,
+      "loss": 2.5355,
+      "step": 355200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.865513928914505e-05,
+      "loss": 2.5141,
+      "step": 355300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.86519372398335e-05,
+      "loss": 2.5175,
+      "step": 355400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.864873519052193e-05,
+      "loss": 2.5303,
+      "step": 355500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.864553314121038e-05,
+      "loss": 2.5186,
+      "step": 355600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.864233109189881e-05,
+      "loss": 2.5079,
+      "step": 355700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.863912904258726e-05,
+      "loss": 2.5176,
+      "step": 355800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.86359269932757e-05,
+      "loss": 2.5181,
+      "step": 355900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.863272494396414e-05,
+      "loss": 2.5136,
+      "step": 356000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.501915216445923,
+      "eval_runtime": 177.034,
+      "eval_samples_per_second": 56.486,
+      "eval_steps_per_second": 3.53,
+      "step": 356000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8629522894652584e-05,
+      "loss": 2.5119,
+      "step": 356100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.862632084534102e-05,
+      "loss": 2.5295,
+      "step": 356200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8623118796029464e-05,
+      "loss": 2.5161,
+      "step": 356300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8619916746717897e-05,
+      "loss": 2.5175,
+      "step": 356400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.861671469740634e-05,
+      "loss": 2.5247,
+      "step": 356500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.861351264809478e-05,
+      "loss": 2.5361,
+      "step": 356600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.861031059878322e-05,
+      "loss": 2.5401,
+      "step": 356700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.860710854947166e-05,
+      "loss": 2.5127,
+      "step": 356800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.86039065001601e-05,
+      "loss": 2.5346,
+      "step": 356900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.860070445084855e-05,
+      "loss": 2.542,
+      "step": 357000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.4995999336242676,
+      "eval_runtime": 180.1686,
+      "eval_samples_per_second": 55.504,
+      "eval_steps_per_second": 3.469,
+      "step": 357000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.859750240153698e-05,
+      "loss": 2.5014,
+      "step": 357100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.859430035222543e-05,
+      "loss": 2.5226,
+      "step": 357200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.859109830291386e-05,
+      "loss": 2.5191,
+      "step": 357300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.858789625360231e-05,
+      "loss": 2.5267,
+      "step": 357400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.858469420429075e-05,
+      "loss": 2.5142,
+      "step": 357500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.858149215497919e-05,
+      "loss": 2.532,
+      "step": 357600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8578290105667633e-05,
+      "loss": 2.5104,
+      "step": 357700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8575088056356066e-05,
+      "loss": 2.5333,
+      "step": 357800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.857188600704451e-05,
+      "loss": 2.5213,
+      "step": 357900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8568683957732946e-05,
+      "loss": 2.5094,
+      "step": 358000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.500349521636963,
+      "eval_runtime": 179.5109,
+      "eval_samples_per_second": 55.707,
+      "eval_steps_per_second": 3.482,
+      "step": 358000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.856548190842139e-05,
+      "loss": 2.5176,
+      "step": 358100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.856227985910983e-05,
+      "loss": 2.5237,
+      "step": 358200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.855907780979827e-05,
+      "loss": 2.5199,
+      "step": 358300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.855587576048671e-05,
+      "loss": 2.5333,
+      "step": 358400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.855267371117515e-05,
+      "loss": 2.5012,
+      "step": 358500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.85494716618636e-05,
+      "loss": 2.5268,
+      "step": 358600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.854626961255203e-05,
+      "loss": 2.5097,
+      "step": 358700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.854306756324048e-05,
+      "loss": 2.5219,
+      "step": 358800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.853986551392892e-05,
+      "loss": 2.5008,
+      "step": 358900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.853666346461736e-05,
+      "loss": 2.5321,
+      "step": 359000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.4990463256835938,
+      "eval_runtime": 176.9791,
+      "eval_samples_per_second": 56.504,
+      "eval_steps_per_second": 3.531,
+      "step": 359000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.85334614153058e-05,
+      "loss": 2.5267,
+      "step": 359100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8530259365994236e-05,
+      "loss": 2.5116,
+      "step": 359200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.852705731668268e-05,
+      "loss": 2.5269,
+      "step": 359300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8523855267371116e-05,
+      "loss": 2.5122,
+      "step": 359400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.852065321805956e-05,
+      "loss": 2.5173,
+      "step": 359500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8517451168748e-05,
+      "loss": 2.5284,
+      "step": 359600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.851424911943644e-05,
+      "loss": 2.5028,
+      "step": 359700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.851104707012488e-05,
+      "loss": 2.5211,
+      "step": 359800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.850784502081332e-05,
+      "loss": 2.515,
+      "step": 359900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.850464297150176e-05,
+      "loss": 2.5045,
+      "step": 360000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.5016562938690186,
+      "eval_runtime": 176.9991,
+      "eval_samples_per_second": 56.497,
+      "eval_steps_per_second": 3.531,
+      "step": 360000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.85014409221902e-05,
+      "loss": 2.5153,
+      "step": 360100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.849823887287865e-05,
+      "loss": 2.4965,
+      "step": 360200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.849503682356708e-05,
+      "loss": 2.516,
+      "step": 360300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.849183477425553e-05,
+      "loss": 2.5196,
+      "step": 360400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8488632724943967e-05,
+      "loss": 2.5213,
+      "step": 360500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8485430675632406e-05,
+      "loss": 2.5319,
+      "step": 360600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8482228626320846e-05,
+      "loss": 2.5034,
+      "step": 360700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8479026577009286e-05,
+      "loss": 2.5259,
+      "step": 360800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.847582452769773e-05,
+      "loss": 2.5339,
+      "step": 360900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8472622478386165e-05,
+      "loss": 2.5079,
+      "step": 361000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.498728036880493,
+      "eval_runtime": 174.5838,
+      "eval_samples_per_second": 57.279,
+      "eval_steps_per_second": 3.58,
+      "step": 361000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.846942042907461e-05,
+      "loss": 2.5167,
+      "step": 361100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.846621837976305e-05,
+      "loss": 2.5128,
+      "step": 361200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.846301633045149e-05,
+      "loss": 2.5308,
+      "step": 361300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.845981428113993e-05,
+      "loss": 2.5099,
+      "step": 361400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.845661223182837e-05,
+      "loss": 2.5277,
+      "step": 361500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.845341018251681e-05,
+      "loss": 2.5191,
+      "step": 361600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.845020813320525e-05,
+      "loss": 2.5065,
+      "step": 361700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.84470060838937e-05,
+      "loss": 2.5314,
+      "step": 361800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8443804034582136e-05,
+      "loss": 2.5064,
+      "step": 361900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8440601985270576e-05,
+      "loss": 2.5229,
+      "step": 362000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.5006937980651855,
+      "eval_runtime": 178.2317,
+      "eval_samples_per_second": 56.107,
+      "eval_steps_per_second": 3.507,
+      "step": 362000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8437399935959016e-05,
+      "loss": 2.5102,
+      "step": 362100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8434197886647456e-05,
+      "loss": 2.5236,
+      "step": 362200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8430995837335895e-05,
+      "loss": 2.4997,
+      "step": 362300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8427793788024335e-05,
+      "loss": 2.5303,
+      "step": 362400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.842459173871278e-05,
+      "loss": 2.503,
+      "step": 362500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8421389689401215e-05,
+      "loss": 2.5286,
+      "step": 362600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.841818764008966e-05,
+      "loss": 2.5163,
+      "step": 362700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.84149855907781e-05,
+      "loss": 2.5096,
+      "step": 362800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.841178354146654e-05,
+      "loss": 2.5302,
+      "step": 362900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.840858149215498e-05,
+      "loss": 2.4782,
+      "step": 363000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.5006227493286133,
+      "eval_runtime": 179.5352,
+      "eval_samples_per_second": 55.699,
+      "eval_steps_per_second": 3.481,
+      "step": 363000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.840537944284342e-05,
+      "loss": 2.5197,
+      "step": 363100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.840217739353186e-05,
+      "loss": 2.4934,
+      "step": 363200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.83989753442203e-05,
+      "loss": 2.5161,
+      "step": 363300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8395773294908746e-05,
+      "loss": 2.4989,
+      "step": 363400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8392571245597186e-05,
+      "loss": 2.5181,
+      "step": 363500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8389369196285626e-05,
+      "loss": 2.5186,
+      "step": 363600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8386167146974065e-05,
+      "loss": 2.4932,
+      "step": 363700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8382965097662505e-05,
+      "loss": 2.5351,
+      "step": 363800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8379763048350945e-05,
+      "loss": 2.5293,
+      "step": 363900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8376560999039384e-05,
+      "loss": 2.5084,
+      "step": 364000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.4995594024658203,
+      "eval_runtime": 178.5309,
+      "eval_samples_per_second": 56.013,
+      "eval_steps_per_second": 3.501,
+      "step": 364000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.837335894972783e-05,
+      "loss": 2.499,
+      "step": 364100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.837015690041627e-05,
+      "loss": 2.4895,
+      "step": 364200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.836695485110471e-05,
+      "loss": 2.5049,
+      "step": 364300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.836375280179315e-05,
+      "loss": 2.515,
+      "step": 364400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.836055075248159e-05,
+      "loss": 2.497,
+      "step": 364500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.835734870317003e-05,
+      "loss": 2.501,
+      "step": 364600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.835414665385847e-05,
+      "loss": 2.4983,
+      "step": 364700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.835094460454691e-05,
+      "loss": 2.513,
+      "step": 364800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.834774255523535e-05,
+      "loss": 2.5129,
+      "step": 364900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8344540505923795e-05,
+      "loss": 2.4935,
+      "step": 365000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.499540090560913,
+      "eval_runtime": 174.256,
+      "eval_samples_per_second": 57.387,
+      "eval_steps_per_second": 3.587,
+      "step": 365000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8341338456612235e-05,
+      "loss": 2.5025,
+      "step": 365100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8338136407300675e-05,
+      "loss": 2.5153,
+      "step": 365200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8334934357989115e-05,
+      "loss": 2.5086,
+      "step": 365300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8331732308677554e-05,
+      "loss": 2.4997,
+      "step": 365400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8328530259365994e-05,
+      "loss": 2.5072,
+      "step": 365500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8325328210054434e-05,
+      "loss": 2.5187,
+      "step": 365600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.832212616074288e-05,
+      "loss": 2.5381,
+      "step": 365700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.831892411143132e-05,
+      "loss": 2.4854,
+      "step": 365800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.831572206211976e-05,
+      "loss": 2.5078,
+      "step": 365900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.83125200128082e-05,
+      "loss": 2.5088,
+      "step": 366000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.4976606369018555,
+      "eval_runtime": 174.5916,
+      "eval_samples_per_second": 57.277,
+      "eval_steps_per_second": 3.58,
+      "step": 366000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.830931796349664e-05,
+      "loss": 2.5265,
+      "step": 366100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.830611591418508e-05,
+      "loss": 2.5259,
+      "step": 366200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.830291386487352e-05,
+      "loss": 2.5045,
+      "step": 366300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.829971181556196e-05,
+      "loss": 2.5072,
+      "step": 366400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8296509766250405e-05,
+      "loss": 2.5038,
+      "step": 366500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8293307716938845e-05,
+      "loss": 2.5233,
+      "step": 366600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8290105667627285e-05,
+      "loss": 2.5389,
+      "step": 366700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8286903618315724e-05,
+      "loss": 2.5118,
+      "step": 366800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8283701569004164e-05,
+      "loss": 2.4953,
+      "step": 366900
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8280499519692604e-05,
+      "loss": 2.5081,
+      "step": 367000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.4992783069610596,
+      "eval_runtime": 174.4627,
+      "eval_samples_per_second": 57.319,
+      "eval_steps_per_second": 3.582,
+      "step": 367000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.8277297470381043e-05,
+      "loss": 2.5201,
+      "step": 367100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.827409542106948e-05,
+      "loss": 2.5032,
+      "step": 367200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.827089337175793e-05,
+      "loss": 2.5098,
+      "step": 367300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.826769132244637e-05,
+      "loss": 2.4947,
+      "step": 367400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.826448927313481e-05,
+      "loss": 2.5109,
+      "step": 367500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.826128722382325e-05,
+      "loss": 2.5234,
+      "step": 367600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.825808517451169e-05,
+      "loss": 2.5035,
+      "step": 367700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.825488312520013e-05,
+      "loss": 2.5126,
+      "step": 367800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.825168107588857e-05,
+      "loss": 2.4972,
+      "step": 367900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.824847902657701e-05,
+      "loss": 2.5145,
+      "step": 368000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.498431444168091,
+      "eval_runtime": 177.0359,
+      "eval_samples_per_second": 56.486,
+      "eval_steps_per_second": 3.53,
+      "step": 368000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8245276977265454e-05,
+      "loss": 2.484,
+      "step": 368100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8242074927953894e-05,
+      "loss": 2.4975,
+      "step": 368200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8238872878642334e-05,
+      "loss": 2.4994,
+      "step": 368300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8235670829330774e-05,
+      "loss": 2.5052,
+      "step": 368400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.823246878001921e-05,
+      "loss": 2.4882,
+      "step": 368500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.822926673070765e-05,
+      "loss": 2.4856,
+      "step": 368600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.822606468139609e-05,
+      "loss": 2.4974,
+      "step": 368700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.822286263208454e-05,
+      "loss": 2.5006,
+      "step": 368800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.821966058277298e-05,
+      "loss": 2.4981,
+      "step": 368900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.821645853346142e-05,
+      "loss": 2.4965,
+      "step": 369000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.499391794204712,
+      "eval_runtime": 176.8888,
+      "eval_samples_per_second": 56.533,
+      "eval_steps_per_second": 3.533,
+      "step": 369000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.821325648414986e-05,
+      "loss": 2.5005,
+      "step": 369100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.82100544348383e-05,
+      "loss": 2.4901,
+      "step": 369200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.820685238552674e-05,
+      "loss": 2.5005,
+      "step": 369300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.820365033621518e-05,
+      "loss": 2.4681,
+      "step": 369400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8200448286903624e-05,
+      "loss": 2.4975,
+      "step": 369500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.819724623759206e-05,
+      "loss": 2.496,
+      "step": 369600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8194044188280504e-05,
+      "loss": 2.5186,
+      "step": 369700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8190842138968944e-05,
+      "loss": 2.4931,
+      "step": 369800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.818764008965738e-05,
+      "loss": 2.4973,
+      "step": 369900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.818443804034582e-05,
+      "loss": 2.5053,
+      "step": 370000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.498286247253418,
+      "eval_runtime": 176.9249,
+      "eval_samples_per_second": 56.521,
+      "eval_steps_per_second": 3.533,
+      "step": 370000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.818123599103426e-05,
+      "loss": 2.4876,
+      "step": 370100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.81780339417227e-05,
+      "loss": 2.4781,
+      "step": 370200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.817483189241114e-05,
+      "loss": 2.512,
+      "step": 370300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.817162984309959e-05,
+      "loss": 2.4961,
+      "step": 370400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.816842779378803e-05,
+      "loss": 2.4942,
+      "step": 370500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.816522574447647e-05,
+      "loss": 2.4921,
+      "step": 370600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.816202369516491e-05,
+      "loss": 2.492,
+      "step": 370700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.815882164585335e-05,
+      "loss": 2.488,
+      "step": 370800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.815561959654179e-05,
+      "loss": 2.4957,
+      "step": 370900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.815241754723023e-05,
+      "loss": 2.4859,
+      "step": 371000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.499032974243164,
+      "eval_runtime": 177.6943,
+      "eval_samples_per_second": 56.276,
+      "eval_steps_per_second": 3.517,
+      "step": 371000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8149215497918674e-05,
+      "loss": 2.5002,
+      "step": 371100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.814601344860711e-05,
+      "loss": 2.4838,
+      "step": 371200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.814281139929555e-05,
+      "loss": 2.4838,
+      "step": 371300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.813960934998399e-05,
+      "loss": 2.477,
+      "step": 371400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.813640730067243e-05,
+      "loss": 2.4773,
+      "step": 371500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.813320525136087e-05,
+      "loss": 2.4868,
+      "step": 371600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.813000320204931e-05,
+      "loss": 2.5113,
+      "step": 371700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.812680115273776e-05,
+      "loss": 2.4901,
+      "step": 371800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.812359910342619e-05,
+      "loss": 2.4744,
+      "step": 371900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.812039705411464e-05,
+      "loss": 2.5048,
+      "step": 372000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.497897148132324,
+      "eval_runtime": 173.8771,
+      "eval_samples_per_second": 57.512,
+      "eval_steps_per_second": 3.594,
+      "step": 372000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.811719500480308e-05,
+      "loss": 2.4887,
+      "step": 372100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.811399295549152e-05,
+      "loss": 2.5085,
+      "step": 372200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.811079090617996e-05,
+      "loss": 2.5021,
+      "step": 372300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.81075888568684e-05,
+      "loss": 2.5068,
+      "step": 372400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.810438680755684e-05,
+      "loss": 2.4976,
+      "step": 372500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8101184758245277e-05,
+      "loss": 2.4909,
+      "step": 372600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.809798270893372e-05,
+      "loss": 2.4874,
+      "step": 372700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8094780659622156e-05,
+      "loss": 2.4913,
+      "step": 372800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.80915786103106e-05,
+      "loss": 2.4973,
+      "step": 372900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.808837656099904e-05,
+      "loss": 2.5019,
+      "step": 373000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.5008938312530518,
+      "eval_runtime": 174.588,
+      "eval_samples_per_second": 57.278,
+      "eval_steps_per_second": 3.58,
+      "step": 373000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.808517451168748e-05,
+      "loss": 2.499,
+      "step": 373100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.808197246237592e-05,
+      "loss": 2.4925,
+      "step": 373200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.807877041306436e-05,
+      "loss": 2.4797,
+      "step": 373300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.807556836375281e-05,
+      "loss": 2.5389,
+      "step": 373400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.807236631444124e-05,
+      "loss": 2.5107,
+      "step": 373500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.806916426512969e-05,
+      "loss": 2.51,
+      "step": 373600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.806596221581813e-05,
+      "loss": 2.5184,
+      "step": 373700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.806276016650657e-05,
+      "loss": 2.4962,
+      "step": 373800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.805955811719501e-05,
+      "loss": 2.522,
+      "step": 373900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8056356067883446e-05,
+      "loss": 2.4978,
+      "step": 374000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.496070146560669,
+      "eval_runtime": 173.6201,
+      "eval_samples_per_second": 57.597,
+      "eval_steps_per_second": 3.6,
+      "step": 374000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.805315401857189e-05,
+      "loss": 2.5088,
+      "step": 374100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8049951969260326e-05,
+      "loss": 2.5204,
+      "step": 374200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.804674991994877e-05,
+      "loss": 2.5002,
+      "step": 374300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8043547870637205e-05,
+      "loss": 2.5399,
+      "step": 374400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.804034582132565e-05,
+      "loss": 2.4997,
+      "step": 374500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.803714377201409e-05,
+      "loss": 2.5279,
+      "step": 374600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.803394172270253e-05,
+      "loss": 2.4982,
+      "step": 374700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.803073967339097e-05,
+      "loss": 2.5392,
+      "step": 374800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.802753762407941e-05,
+      "loss": 2.5068,
+      "step": 374900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.802433557476786e-05,
+      "loss": 2.5149,
+      "step": 375000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.4967455863952637,
+      "eval_runtime": 176.9338,
+      "eval_samples_per_second": 56.518,
+      "eval_steps_per_second": 3.532,
+      "step": 375000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.802113352545629e-05,
+      "loss": 2.5,
+      "step": 375100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.801793147614474e-05,
+      "loss": 2.5314,
+      "step": 375200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8014729426833177e-05,
+      "loss": 2.5022,
+      "step": 375300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8011527377521616e-05,
+      "loss": 2.5078,
+      "step": 375400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8008325328210056e-05,
+      "loss": 2.4933,
+      "step": 375500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8005123278898496e-05,
+      "loss": 2.5119,
+      "step": 375600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.800192122958694e-05,
+      "loss": 2.4983,
+      "step": 375700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7998719180275375e-05,
+      "loss": 2.4915,
+      "step": 375800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.799551713096382e-05,
+      "loss": 2.4841,
+      "step": 375900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7992315081652255e-05,
+      "loss": 2.5027,
+      "step": 376000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.497227430343628,
+      "eval_runtime": 177.9629,
+      "eval_samples_per_second": 56.191,
+      "eval_steps_per_second": 3.512,
+      "step": 376000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.79891130323407e-05,
+      "loss": 2.5291,
+      "step": 376100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.798591098302914e-05,
+      "loss": 2.519,
+      "step": 376200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.798270893371758e-05,
+      "loss": 2.504,
+      "step": 376300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.797950688440603e-05,
+      "loss": 2.5066,
+      "step": 376400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.797630483509446e-05,
+      "loss": 2.5211,
+      "step": 376500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.797310278578291e-05,
+      "loss": 2.5055,
+      "step": 376600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.796990073647134e-05,
+      "loss": 2.512,
+      "step": 376700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7966698687159786e-05,
+      "loss": 2.4845,
+      "step": 376800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7963496637848226e-05,
+      "loss": 2.5076,
+      "step": 376900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7960294588536666e-05,
+      "loss": 2.507,
+      "step": 377000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.4986350536346436,
+      "eval_runtime": 178.4737,
+      "eval_samples_per_second": 56.031,
+      "eval_steps_per_second": 3.502,
+      "step": 377000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7957092539225105e-05,
+      "loss": 2.5007,
+      "step": 377100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7953890489913545e-05,
+      "loss": 2.5171,
+      "step": 377200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.795068844060199e-05,
+      "loss": 2.5309,
+      "step": 377300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7947486391290425e-05,
+      "loss": 2.5166,
+      "step": 377400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.794428434197887e-05,
+      "loss": 2.4914,
+      "step": 377500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7941082292667304e-05,
+      "loss": 2.5082,
+      "step": 377600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.793788024335575e-05,
+      "loss": 2.5063,
+      "step": 377700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.793467819404419e-05,
+      "loss": 2.5043,
+      "step": 377800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.793147614473263e-05,
+      "loss": 2.508,
+      "step": 377900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.792827409542108e-05,
+      "loss": 2.5115,
+      "step": 378000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.497148036956787,
+      "eval_runtime": 177.947,
+      "eval_samples_per_second": 56.196,
+      "eval_steps_per_second": 3.512,
+      "step": 378000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.792507204610951e-05,
+      "loss": 2.5246,
+      "step": 378100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7921869996797956e-05,
+      "loss": 2.5,
+      "step": 378200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.791866794748639e-05,
+      "loss": 2.4988,
+      "step": 378300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7915465898174836e-05,
+      "loss": 2.486,
+      "step": 378400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7912263848863275e-05,
+      "loss": 2.5065,
+      "step": 378500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7909061799551715e-05,
+      "loss": 2.5117,
+      "step": 378600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7905859750240155e-05,
+      "loss": 2.512,
+      "step": 378700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7902657700928595e-05,
+      "loss": 2.5151,
+      "step": 378800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.789945565161704e-05,
+      "loss": 2.4957,
+      "step": 378900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7896253602305474e-05,
+      "loss": 2.5088,
+      "step": 379000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.4963321685791016,
+      "eval_runtime": 174.046,
+      "eval_samples_per_second": 57.456,
+      "eval_steps_per_second": 3.591,
+      "step": 379000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.789305155299392e-05,
+      "loss": 2.5089,
+      "step": 379100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7889849503682353e-05,
+      "loss": 2.5041,
+      "step": 379200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.78866474543708e-05,
+      "loss": 2.5202,
+      "step": 379300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.788344540505924e-05,
+      "loss": 2.4987,
+      "step": 379400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.788024335574768e-05,
+      "loss": 2.4961,
+      "step": 379500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7877041306436126e-05,
+      "loss": 2.509,
+      "step": 379600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.787383925712456e-05,
+      "loss": 2.4977,
+      "step": 379700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7870637207813005e-05,
+      "loss": 2.5113,
+      "step": 379800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.786743515850144e-05,
+      "loss": 2.5187,
+      "step": 379900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7864233109189885e-05,
+      "loss": 2.491,
+      "step": 380000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.4965128898620605,
+      "eval_runtime": 173.6496,
+      "eval_samples_per_second": 57.587,
+      "eval_steps_per_second": 3.599,
+      "step": 380000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7861031059878325e-05,
+      "loss": 2.5134,
+      "step": 380100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7857829010566764e-05,
+      "loss": 2.5061,
+      "step": 380200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7854626961255204e-05,
+      "loss": 2.5033,
+      "step": 380300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7851424911943644e-05,
+      "loss": 2.5135,
+      "step": 380400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.784822286263209e-05,
+      "loss": 2.5063,
+      "step": 380500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7845020813320523e-05,
+      "loss": 2.521,
+      "step": 380600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.784181876400897e-05,
+      "loss": 2.5078,
+      "step": 380700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.78386167146974e-05,
+      "loss": 2.5084,
+      "step": 380800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.783541466538585e-05,
+      "loss": 2.5144,
+      "step": 380900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.783221261607429e-05,
+      "loss": 2.4836,
+      "step": 381000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.4974985122680664,
+      "eval_runtime": 174.9398,
+      "eval_samples_per_second": 57.163,
+      "eval_steps_per_second": 3.573,
+      "step": 381000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.782901056676273e-05,
+      "loss": 2.5083,
+      "step": 381100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7825808517451175e-05,
+      "loss": 2.5283,
+      "step": 381200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.782260646813961e-05,
+      "loss": 2.5143,
+      "step": 381300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7819404418828055e-05,
+      "loss": 2.524,
+      "step": 381400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.781620236951649e-05,
+      "loss": 2.5112,
+      "step": 381500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7813000320204934e-05,
+      "loss": 2.516,
+      "step": 381600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7809798270893374e-05,
+      "loss": 2.4911,
+      "step": 381700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7806596221581814e-05,
+      "loss": 2.5095,
+      "step": 381800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7803394172270254e-05,
+      "loss": 2.5176,
+      "step": 381900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.780019212295869e-05,
+      "loss": 2.501,
+      "step": 382000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.49737548828125,
+      "eval_runtime": 176.7872,
+      "eval_samples_per_second": 56.565,
+      "eval_steps_per_second": 3.535,
+      "step": 382000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.779699007364714e-05,
+      "loss": 2.515,
+      "step": 382100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.779378802433557e-05,
+      "loss": 2.5137,
+      "step": 382200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.779058597502402e-05,
+      "loss": 2.4989,
+      "step": 382300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.778738392571245e-05,
+      "loss": 2.5079,
+      "step": 382400
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.77841818764009e-05,
+      "loss": 2.5006,
+      "step": 382500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.778097982708934e-05,
+      "loss": 2.5061,
+      "step": 382600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 2.5081,
+      "step": 382700
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.7774575728466225e-05,
+      "loss": 2.5005,
+      "step": 382800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.777137367915466e-05,
+      "loss": 2.4997,
+      "step": 382900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7768171629843104e-05,
+      "loss": 2.5135,
+      "step": 383000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.496156692504883,
+      "eval_runtime": 177.1933,
+      "eval_samples_per_second": 56.436,
+      "eval_steps_per_second": 3.527,
+      "step": 383000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.776496958053154e-05,
+      "loss": 2.4954,
+      "step": 383100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7761767531219984e-05,
+      "loss": 2.5033,
+      "step": 383200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7758565481908423e-05,
+      "loss": 2.5158,
+      "step": 383300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.775536343259686e-05,
+      "loss": 2.4833,
+      "step": 383400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.77521613832853e-05,
+      "loss": 2.5069,
+      "step": 383500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.774895933397374e-05,
+      "loss": 2.5084,
+      "step": 383600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.774575728466219e-05,
+      "loss": 2.5297,
+      "step": 383700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.774255523535062e-05,
+      "loss": 2.5019,
+      "step": 383800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.773935318603907e-05,
+      "loss": 2.5263,
+      "step": 383900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.773615113672751e-05,
+      "loss": 2.5243,
+      "step": 384000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.49690318107605,
+      "eval_runtime": 176.5316,
+      "eval_samples_per_second": 56.647,
+      "eval_steps_per_second": 3.54,
+      "step": 384000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.773294908741595e-05,
+      "loss": 2.5084,
+      "step": 384100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.772974703810439e-05,
+      "loss": 2.5146,
+      "step": 384200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.772654498879283e-05,
+      "loss": 2.5167,
+      "step": 384300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7723342939481274e-05,
+      "loss": 2.5029,
+      "step": 384400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.772014089016971e-05,
+      "loss": 2.4988,
+      "step": 384500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7716938840858154e-05,
+      "loss": 2.5149,
+      "step": 384600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7713736791546587e-05,
+      "loss": 2.4904,
+      "step": 384700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.771053474223503e-05,
+      "loss": 2.508,
+      "step": 384800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.770733269292347e-05,
+      "loss": 2.511,
+      "step": 384900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.770413064361191e-05,
+      "loss": 2.518,
+      "step": 385000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.496809482574463,
+      "eval_runtime": 177.0211,
+      "eval_samples_per_second": 56.49,
+      "eval_steps_per_second": 3.531,
+      "step": 385000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.770092859430035e-05,
+      "loss": 2.513,
+      "step": 385100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.769772654498879e-05,
+      "loss": 2.4858,
+      "step": 385200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.769452449567724e-05,
+      "loss": 2.515,
+      "step": 385300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.769132244636567e-05,
+      "loss": 2.5217,
+      "step": 385400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.768812039705412e-05,
+      "loss": 2.5021,
+      "step": 385500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.768491834774256e-05,
+      "loss": 2.4919,
+      "step": 385600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7681716298431e-05,
+      "loss": 2.4961,
+      "step": 385700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.767851424911944e-05,
+      "loss": 2.5035,
+      "step": 385800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.767531219980788e-05,
+      "loss": 2.5346,
+      "step": 385900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7672110150496323e-05,
+      "loss": 2.5015,
+      "step": 386000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.4960687160491943,
+      "eval_runtime": 177.1294,
+      "eval_samples_per_second": 56.456,
+      "eval_steps_per_second": 3.528,
+      "step": 386000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7668908101184756e-05,
+      "loss": 2.4868,
+      "step": 386100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.76657060518732e-05,
+      "loss": 2.5166,
+      "step": 386200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.766250400256164e-05,
+      "loss": 2.4984,
+      "step": 386300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.765930195325008e-05,
+      "loss": 2.493,
+      "step": 386400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.765609990393852e-05,
+      "loss": 2.4937,
+      "step": 386500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.765289785462696e-05,
+      "loss": 2.5139,
+      "step": 386600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.76496958053154e-05,
+      "loss": 2.4961,
+      "step": 386700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.764649375600384e-05,
+      "loss": 2.5078,
+      "step": 386800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.764329170669229e-05,
+      "loss": 2.5074,
+      "step": 386900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.764008965738073e-05,
+      "loss": 2.5123,
+      "step": 387000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.496873378753662,
+      "eval_runtime": 173.3967,
+      "eval_samples_per_second": 57.671,
+      "eval_steps_per_second": 3.604,
+      "step": 387000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.763688760806917e-05,
+      "loss": 2.492,
+      "step": 387100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.763368555875761e-05,
+      "loss": 2.5076,
+      "step": 387200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.763048350944605e-05,
+      "loss": 2.5064,
+      "step": 387300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7627281460134487e-05,
+      "loss": 2.5212,
+      "step": 387400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7624079410822926e-05,
+      "loss": 2.5205,
+      "step": 387500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.762087736151137e-05,
+      "loss": 2.4969,
+      "step": 387600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7617675312199806e-05,
+      "loss": 2.4872,
+      "step": 387700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.761447326288825e-05,
+      "loss": 2.4822,
+      "step": 387800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.761127121357669e-05,
+      "loss": 2.4989,
+      "step": 387900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.760806916426513e-05,
+      "loss": 2.5161,
+      "step": 388000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.4972026348114014,
+      "eval_runtime": 173.6832,
+      "eval_samples_per_second": 57.576,
+      "eval_steps_per_second": 3.599,
+      "step": 388000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.760486711495357e-05,
+      "loss": 2.5165,
+      "step": 388100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.760166506564201e-05,
+      "loss": 2.4943,
+      "step": 388200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.759846301633045e-05,
+      "loss": 2.4975,
+      "step": 388300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.759526096701889e-05,
+      "loss": 2.5096,
+      "step": 388400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.759205891770734e-05,
+      "loss": 2.4847,
+      "step": 388500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.758885686839578e-05,
+      "loss": 2.4915,
+      "step": 388600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.758565481908422e-05,
+      "loss": 2.5114,
+      "step": 388700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7582452769772657e-05,
+      "loss": 2.495,
+      "step": 388800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7579250720461096e-05,
+      "loss": 2.5039,
+      "step": 388900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7576048671149536e-05,
+      "loss": 2.4924,
+      "step": 389000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.4967758655548096,
+      "eval_runtime": 174.0389,
+      "eval_samples_per_second": 57.458,
+      "eval_steps_per_second": 3.591,
+      "step": 389000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7572846621837976e-05,
+      "loss": 2.4837,
+      "step": 389100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.756964457252642e-05,
+      "loss": 2.4728,
+      "step": 389200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.756644252321486e-05,
+      "loss": 2.4844,
+      "step": 389300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.75632404739033e-05,
+      "loss": 2.5008,
+      "step": 389400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.756003842459174e-05,
+      "loss": 2.4834,
+      "step": 389500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.755683637528018e-05,
+      "loss": 2.4868,
+      "step": 389600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.755363432596862e-05,
+      "loss": 2.4877,
+      "step": 389700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.755043227665706e-05,
+      "loss": 2.4949,
+      "step": 389800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.75472302273455e-05,
+      "loss": 2.5034,
+      "step": 389900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.754402817803394e-05,
+      "loss": 2.4923,
+      "step": 390000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.496317148208618,
+      "eval_runtime": 173.3618,
+      "eval_samples_per_second": 57.683,
+      "eval_steps_per_second": 3.605,
+      "step": 390000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.754082612872239e-05,
+      "loss": 2.4867,
+      "step": 390100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7537624079410826e-05,
+      "loss": 2.5088,
+      "step": 390200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7534422030099266e-05,
+      "loss": 2.4966,
+      "step": 390300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7531219980787706e-05,
+      "loss": 2.488,
+      "step": 390400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7528017931476146e-05,
+      "loss": 2.5066,
+      "step": 390500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7524815882164585e-05,
+      "loss": 2.4949,
+      "step": 390600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7521613832853025e-05,
+      "loss": 2.5074,
+      "step": 390700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.751841178354147e-05,
+      "loss": 2.5273,
+      "step": 390800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.751520973422991e-05,
+      "loss": 2.5095,
+      "step": 390900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.751200768491835e-05,
+      "loss": 2.4876,
+      "step": 391000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.4965176582336426,
+      "eval_runtime": 173.8711,
+      "eval_samples_per_second": 57.514,
+      "eval_steps_per_second": 3.595,
+      "step": 391000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.750880563560679e-05,
+      "loss": 2.4904,
+      "step": 391100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.750560358629523e-05,
+      "loss": 2.5088,
+      "step": 391200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.750240153698367e-05,
+      "loss": 2.4968,
+      "step": 391300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.749919948767211e-05,
+      "loss": 2.4927,
+      "step": 391400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.749599743836055e-05,
+      "loss": 2.4834,
+      "step": 391500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7492795389048996e-05,
+      "loss": 2.5241,
+      "step": 391600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7489593339737436e-05,
+      "loss": 2.4944,
+      "step": 391700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7486391290425876e-05,
+      "loss": 2.4965,
+      "step": 391800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7483189241114316e-05,
+      "loss": 2.4864,
+      "step": 391900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7479987191802755e-05,
+      "loss": 2.4961,
+      "step": 392000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.496955156326294,
+      "eval_runtime": 177.1413,
+      "eval_samples_per_second": 56.452,
+      "eval_steps_per_second": 3.528,
+      "step": 392000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7476785142491195e-05,
+      "loss": 2.4987,
+      "step": 392100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7473583093179635e-05,
+      "loss": 2.5002,
+      "step": 392200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7470381043868074e-05,
+      "loss": 2.4853,
+      "step": 392300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.746717899455652e-05,
+      "loss": 2.5047,
+      "step": 392400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.746397694524496e-05,
+      "loss": 2.4739,
+      "step": 392500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.74607748959334e-05,
+      "loss": 2.4956,
+      "step": 392600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.745757284662184e-05,
+      "loss": 2.5093,
+      "step": 392700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.745437079731028e-05,
+      "loss": 2.487,
+      "step": 392800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.745116874799872e-05,
+      "loss": 2.5049,
+      "step": 392900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.744796669868716e-05,
+      "loss": 2.4893,
+      "step": 393000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.497549533843994,
+      "eval_runtime": 177.5778,
+      "eval_samples_per_second": 56.313,
+      "eval_steps_per_second": 3.52,
+      "step": 393000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.74447646493756e-05,
+      "loss": 2.501,
+      "step": 393100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7441562600064046e-05,
+      "loss": 2.4826,
+      "step": 393200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7438360550752485e-05,
+      "loss": 2.4819,
+      "step": 393300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7435158501440925e-05,
+      "loss": 2.4699,
+      "step": 393400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7431956452129365e-05,
+      "loss": 2.5042,
+      "step": 393500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7428754402817805e-05,
+      "loss": 2.5126,
+      "step": 393600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7425552353506244e-05,
+      "loss": 2.5198,
+      "step": 393700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7422350304194684e-05,
+      "loss": 2.5092,
+      "step": 393800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.741914825488313e-05,
+      "loss": 2.5063,
+      "step": 393900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.741594620557157e-05,
+      "loss": 2.4989,
+      "step": 394000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.497384548187256,
+      "eval_runtime": 178.2295,
+      "eval_samples_per_second": 56.107,
+      "eval_steps_per_second": 3.507,
+      "step": 394000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.741274415626001e-05,
+      "loss": 2.4731,
+      "step": 394100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.740954210694845e-05,
+      "loss": 2.4909,
+      "step": 394200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.740634005763689e-05,
+      "loss": 2.4864,
+      "step": 394300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.740313800832533e-05,
+      "loss": 2.488,
+      "step": 394400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.739993595901377e-05,
+      "loss": 2.4725,
+      "step": 394500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.739673390970221e-05,
+      "loss": 2.482,
+      "step": 394600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.739353186039065e-05,
+      "loss": 2.4969,
+      "step": 394700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7390329811079095e-05,
+      "loss": 2.4944,
+      "step": 394800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7387127761767535e-05,
+      "loss": 2.5095,
+      "step": 394900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7383925712455975e-05,
+      "loss": 2.4977,
+      "step": 395000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.4954965114593506,
+      "eval_runtime": 176.1394,
+      "eval_samples_per_second": 56.773,
+      "eval_steps_per_second": 3.548,
+      "step": 395000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7380723663144414e-05,
+      "loss": 2.4841,
+      "step": 395100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7377521613832854e-05,
+      "loss": 2.4834,
+      "step": 395200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7374319564521294e-05,
+      "loss": 2.5135,
+      "step": 395300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7371117515209733e-05,
+      "loss": 2.4807,
+      "step": 395400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.736791546589818e-05,
+      "loss": 2.4932,
+      "step": 395500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.736471341658662e-05,
+      "loss": 2.4871,
+      "step": 395600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.736151136727506e-05,
+      "loss": 2.4998,
+      "step": 395700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.73583093179635e-05,
+      "loss": 2.4979,
+      "step": 395800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.735510726865194e-05,
+      "loss": 2.4878,
+      "step": 395900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.735190521934038e-05,
+      "loss": 2.4836,
+      "step": 396000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.49569034576416,
+      "eval_runtime": 176.8797,
+      "eval_samples_per_second": 56.536,
+      "eval_steps_per_second": 3.533,
+      "step": 396000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.734870317002882e-05,
+      "loss": 2.4959,
+      "step": 396100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7345501120717265e-05,
+      "loss": 2.4945,
+      "step": 396200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.73422990714057e-05,
+      "loss": 2.4926,
+      "step": 396300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7339097022094144e-05,
+      "loss": 2.4823,
+      "step": 396400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7335894972782584e-05,
+      "loss": 2.498,
+      "step": 396500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7332692923471024e-05,
+      "loss": 2.5038,
+      "step": 396600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7329490874159464e-05,
+      "loss": 2.4799,
+      "step": 396700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.73262888248479e-05,
+      "loss": 2.4956,
+      "step": 396800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.732308677553635e-05,
+      "loss": 2.4837,
+      "step": 396900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.731988472622478e-05,
+      "loss": 2.4958,
+      "step": 397000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.49448299407959,
+      "eval_runtime": 174.5573,
+      "eval_samples_per_second": 57.288,
+      "eval_steps_per_second": 3.58,
+      "step": 397000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.731668267691323e-05,
+      "loss": 2.493,
+      "step": 397100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.731348062760167e-05,
+      "loss": 2.4892,
+      "step": 397200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.731027857829011e-05,
+      "loss": 2.4825,
+      "step": 397300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.730707652897855e-05,
+      "loss": 2.4979,
+      "step": 397400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.730387447966699e-05,
+      "loss": 2.509,
+      "step": 397500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.730067243035543e-05,
+      "loss": 2.5102,
+      "step": 397600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.729747038104387e-05,
+      "loss": 2.5075,
+      "step": 397700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7294268331732314e-05,
+      "loss": 2.4701,
+      "step": 397800
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.729106628242075e-05,
+      "loss": 2.4921,
+      "step": 397900
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7287864233109194e-05,
+      "loss": 2.4906,
+      "step": 398000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.4942800998687744,
+      "eval_runtime": 173.664,
+      "eval_samples_per_second": 57.582,
+      "eval_steps_per_second": 3.599,
+      "step": 398000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.7284662183797634e-05,
+      "loss": 2.4884,
+      "step": 398100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.728146013448607e-05,
+      "loss": 2.5008,
+      "step": 398200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.727825808517451e-05,
+      "loss": 2.4798,
+      "step": 398300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.727505603586295e-05,
+      "loss": 2.4974,
+      "step": 398400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.72718539865514e-05,
+      "loss": 2.4839,
+      "step": 398500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.726865193723983e-05,
+      "loss": 2.4923,
+      "step": 398600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.726544988792828e-05,
+      "loss": 2.5026,
+      "step": 398700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.726224783861672e-05,
+      "loss": 2.4937,
+      "step": 398800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.725904578930516e-05,
+      "loss": 2.4812,
+      "step": 398900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.72558437399936e-05,
+      "loss": 2.5094,
+      "step": 399000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.4958760738372803,
+      "eval_runtime": 173.8851,
+      "eval_samples_per_second": 57.509,
+      "eval_steps_per_second": 3.594,
+      "step": 399000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.725264169068204e-05,
+      "loss": 2.4947,
+      "step": 399100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7249439641370484e-05,
+      "loss": 2.4721,
+      "step": 399200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.724623759205892e-05,
+      "loss": 2.4907,
+      "step": 399300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7243035542747364e-05,
+      "loss": 2.5079,
+      "step": 399400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.72398334934358e-05,
+      "loss": 2.5069,
+      "step": 399500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.723663144412424e-05,
+      "loss": 2.4893,
+      "step": 399600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.723342939481268e-05,
+      "loss": 2.5025,
+      "step": 399700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.723022734550112e-05,
+      "loss": 2.4968,
+      "step": 399800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.722702529618956e-05,
+      "loss": 2.4965,
+      "step": 399900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7223823246878e-05,
+      "loss": 2.4971,
+      "step": 400000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.4941930770874023,
+      "eval_runtime": 174.0267,
+      "eval_samples_per_second": 57.462,
+      "eval_steps_per_second": 3.591,
+      "step": 400000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.722062119756645e-05,
+      "loss": 2.4825,
+      "step": 400100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.721741914825488e-05,
+      "loss": 2.4765,
+      "step": 400200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.721421709894333e-05,
+      "loss": 2.4687,
+      "step": 400300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.721101504963176e-05,
+      "loss": 2.501,
+      "step": 400400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.720781300032021e-05,
+      "loss": 2.4899,
+      "step": 400500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.720461095100865e-05,
+      "loss": 2.4968,
+      "step": 400600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.720140890169709e-05,
+      "loss": 2.4797,
+      "step": 400700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7198206852385534e-05,
+      "loss": 2.4911,
+      "step": 400800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7195004803073967e-05,
+      "loss": 2.4909,
+      "step": 400900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.719180275376241e-05,
+      "loss": 2.5164,
+      "step": 401000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.494380474090576,
+      "eval_runtime": 174.0163,
+      "eval_samples_per_second": 57.466,
+      "eval_steps_per_second": 3.592,
+      "step": 401000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7188600704450846e-05,
+      "loss": 2.4886,
+      "step": 401100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.718539865513929e-05,
+      "loss": 2.4859,
+      "step": 401200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.718219660582773e-05,
+      "loss": 2.5117,
+      "step": 401300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.717899455651617e-05,
+      "loss": 2.4978,
+      "step": 401400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.717579250720462e-05,
+      "loss": 2.4842,
+      "step": 401500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.717259045789305e-05,
+      "loss": 2.5005,
+      "step": 401600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.71693884085815e-05,
+      "loss": 2.5048,
+      "step": 401700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.716618635926993e-05,
+      "loss": 2.5005,
+      "step": 401800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.716298430995838e-05,
+      "loss": 2.5101,
+      "step": 401900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.715978226064681e-05,
+      "loss": 2.4978,
+      "step": 402000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.4960920810699463,
+      "eval_runtime": 177.2252,
+      "eval_samples_per_second": 56.425,
+      "eval_steps_per_second": 3.527,
+      "step": 402000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.715658021133526e-05,
+      "loss": 2.506,
+      "step": 402100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.71533781620237e-05,
+      "loss": 2.5091,
+      "step": 402200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7150176112712136e-05,
+      "loss": 2.4909,
+      "step": 402300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.714697406340058e-05,
+      "loss": 2.4901,
+      "step": 402400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7143772014089016e-05,
+      "loss": 2.4949,
+      "step": 402500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.714056996477746e-05,
+      "loss": 2.4997,
+      "step": 402600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7137367915465895e-05,
+      "loss": 2.5078,
+      "step": 402700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.713416586615434e-05,
+      "loss": 2.4992,
+      "step": 402800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.713096381684278e-05,
+      "loss": 2.5028,
+      "step": 402900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.712776176753122e-05,
+      "loss": 2.4893,
+      "step": 403000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.495441198348999,
+      "eval_runtime": 176.8442,
+      "eval_samples_per_second": 56.547,
+      "eval_steps_per_second": 3.534,
+      "step": 403000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.712455971821967e-05,
+      "loss": 2.4648,
+      "step": 403100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.71213576689081e-05,
+      "loss": 2.5059,
+      "step": 403200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.711815561959655e-05,
+      "loss": 2.4918,
+      "step": 403300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.711495357028498e-05,
+      "loss": 2.5006,
+      "step": 403400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.711175152097343e-05,
+      "loss": 2.48,
+      "step": 403500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.710854947166186e-05,
+      "loss": 2.4921,
+      "step": 403600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7105347422350306e-05,
+      "loss": 2.49,
+      "step": 403700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7102145373038746e-05,
+      "loss": 2.4854,
+      "step": 403800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7098943323727186e-05,
+      "loss": 2.4964,
+      "step": 403900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.709574127441563e-05,
+      "loss": 2.4902,
+      "step": 404000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.49703311920166,
+      "eval_runtime": 176.7332,
+      "eval_samples_per_second": 56.582,
+      "eval_steps_per_second": 3.536,
+      "step": 404000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7092539225104065e-05,
+      "loss": 2.4859,
+      "step": 404100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.708933717579251e-05,
+      "loss": 2.493,
+      "step": 404200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7086135126480945e-05,
+      "loss": 2.5139,
+      "step": 404300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.708293307716939e-05,
+      "loss": 2.4837,
+      "step": 404400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.707973102785783e-05,
+      "loss": 2.5094,
+      "step": 404500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.707652897854627e-05,
+      "loss": 2.4963,
+      "step": 404600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.707332692923472e-05,
+      "loss": 2.4897,
+      "step": 404700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.707012487992315e-05,
+      "loss": 2.5022,
+      "step": 404800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.70669228306116e-05,
+      "loss": 2.4878,
+      "step": 404900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.706372078130003e-05,
+      "loss": 2.5035,
+      "step": 405000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.496225595474243,
+      "eval_runtime": 177.3895,
+      "eval_samples_per_second": 56.373,
+      "eval_steps_per_second": 3.523,
+      "step": 405000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7060518731988476e-05,
+      "loss": 2.4765,
+      "step": 405100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.705731668267691e-05,
+      "loss": 2.5169,
+      "step": 405200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7054114633365356e-05,
+      "loss": 2.4945,
+      "step": 405300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7050912584053795e-05,
+      "loss": 2.4851,
+      "step": 405400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7047710534742235e-05,
+      "loss": 2.4956,
+      "step": 405500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.704450848543068e-05,
+      "loss": 2.4983,
+      "step": 405600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7041306436119115e-05,
+      "loss": 2.4676,
+      "step": 405700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.703810438680756e-05,
+      "loss": 2.4784,
+      "step": 405800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7034902337495994e-05,
+      "loss": 2.4799,
+      "step": 405900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.703170028818444e-05,
+      "loss": 2.4875,
+      "step": 406000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.493391752243042,
+      "eval_runtime": 176.6539,
+      "eval_samples_per_second": 56.608,
+      "eval_steps_per_second": 3.538,
+      "step": 406000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.702849823887288e-05,
+      "loss": 2.484,
+      "step": 406100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.702529618956132e-05,
+      "loss": 2.4861,
+      "step": 406200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.702209414024977e-05,
+      "loss": 2.479,
+      "step": 406300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.70188920909382e-05,
+      "loss": 2.497,
+      "step": 406400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7015690041626646e-05,
+      "loss": 2.4992,
+      "step": 406500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.701248799231508e-05,
+      "loss": 2.4795,
+      "step": 406600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7009285943003526e-05,
+      "loss": 2.5039,
+      "step": 406700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7006083893691965e-05,
+      "loss": 2.4807,
+      "step": 406800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7002881844380405e-05,
+      "loss": 2.4756,
+      "step": 406900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6999679795068845e-05,
+      "loss": 2.5064,
+      "step": 407000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.4955644607543945,
+      "eval_runtime": 176.7613,
+      "eval_samples_per_second": 56.573,
+      "eval_steps_per_second": 3.536,
+      "step": 407000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6996477745757285e-05,
+      "loss": 2.4904,
+      "step": 407100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.699327569644573e-05,
+      "loss": 2.483,
+      "step": 407200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6990073647134164e-05,
+      "loss": 2.473,
+      "step": 407300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.698687159782261e-05,
+      "loss": 2.4728,
+      "step": 407400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6983669548511043e-05,
+      "loss": 2.4798,
+      "step": 407500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.698046749919949e-05,
+      "loss": 2.495,
+      "step": 407600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.697726544988793e-05,
+      "loss": 2.4861,
+      "step": 407700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.697406340057637e-05,
+      "loss": 2.4784,
+      "step": 407800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6970861351264816e-05,
+      "loss": 2.4936,
+      "step": 407900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.696765930195325e-05,
+      "loss": 2.4981,
+      "step": 408000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.494189739227295,
+      "eval_runtime": 177.0105,
+      "eval_samples_per_second": 56.494,
+      "eval_steps_per_second": 3.531,
+      "step": 408000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6964457252641695e-05,
+      "loss": 2.4751,
+      "step": 408100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.696125520333013e-05,
+      "loss": 2.4802,
+      "step": 408200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6958053154018575e-05,
+      "loss": 2.4841,
+      "step": 408300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6954851104707015e-05,
+      "loss": 2.4799,
+      "step": 408400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6951649055395454e-05,
+      "loss": 2.4673,
+      "step": 408500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6948447006083894e-05,
+      "loss": 2.4727,
+      "step": 408600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6945244956772334e-05,
+      "loss": 2.4801,
+      "step": 408700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.694204290746078e-05,
+      "loss": 2.4717,
+      "step": 408800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.693884085814921e-05,
+      "loss": 2.4838,
+      "step": 408900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.693563880883766e-05,
+      "loss": 2.4908,
+      "step": 409000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.4922473430633545,
+      "eval_runtime": 177.4522,
+      "eval_samples_per_second": 56.353,
+      "eval_steps_per_second": 3.522,
+      "step": 409000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.69324367595261e-05,
+      "loss": 2.4793,
+      "step": 409100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.692923471021454e-05,
+      "loss": 2.4808,
+      "step": 409200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.692603266090298e-05,
+      "loss": 2.4761,
+      "step": 409300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.692283061159142e-05,
+      "loss": 2.469,
+      "step": 409400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6919628562279865e-05,
+      "loss": 2.4551,
+      "step": 409500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.69164265129683e-05,
+      "loss": 2.5,
+      "step": 409600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6913224463656745e-05,
+      "loss": 2.4481,
+      "step": 409700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.691002241434518e-05,
+      "loss": 2.4726,
+      "step": 409800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6906820365033624e-05,
+      "loss": 2.474,
+      "step": 409900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6903618315722064e-05,
+      "loss": 2.4847,
+      "step": 410000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.49210262298584,
+      "eval_runtime": 174.6081,
+      "eval_samples_per_second": 57.271,
+      "eval_steps_per_second": 3.579,
+      "step": 410000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6900416266410504e-05,
+      "loss": 2.4922,
+      "step": 410100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6897214217098944e-05,
+      "loss": 2.4802,
+      "step": 410200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.689401216778738e-05,
+      "loss": 2.4912,
+      "step": 410300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.689081011847583e-05,
+      "loss": 2.4649,
+      "step": 410400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.688760806916426e-05,
+      "loss": 2.4839,
+      "step": 410500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.688440601985271e-05,
+      "loss": 2.4636,
+      "step": 410600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.688120397054115e-05,
+      "loss": 2.4508,
+      "step": 410700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.687800192122959e-05,
+      "loss": 2.4876,
+      "step": 410800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.687479987191803e-05,
+      "loss": 2.4915,
+      "step": 410900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.687159782260647e-05,
+      "loss": 2.477,
+      "step": 411000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.4932756423950195,
+      "eval_runtime": 174.1704,
+      "eval_samples_per_second": 57.415,
+      "eval_steps_per_second": 3.588,
+      "step": 411000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6868395773294915e-05,
+      "loss": 2.4905,
+      "step": 411100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.686519372398335e-05,
+      "loss": 2.4914,
+      "step": 411200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6861991674671794e-05,
+      "loss": 2.4911,
+      "step": 411300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6858789625360234e-05,
+      "loss": 2.509,
+      "step": 411400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6855587576048674e-05,
+      "loss": 2.4944,
+      "step": 411500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6852385526737113e-05,
+      "loss": 2.4959,
+      "step": 411600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.684918347742555e-05,
+      "loss": 2.4756,
+      "step": 411700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.684598142811399e-05,
+      "loss": 2.4881,
+      "step": 411800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.684277937880243e-05,
+      "loss": 2.5099,
+      "step": 411900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.683957732949088e-05,
+      "loss": 2.5014,
+      "step": 412000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.495574712753296,
+      "eval_runtime": 173.897,
+      "eval_samples_per_second": 57.505,
+      "eval_steps_per_second": 3.594,
+      "step": 412000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.683637528017931e-05,
+      "loss": 2.4868,
+      "step": 412100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.683317323086776e-05,
+      "loss": 2.4963,
+      "step": 412200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.68299711815562e-05,
+      "loss": 2.4964,
+      "step": 412300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.682676913224464e-05,
+      "loss": 2.4912,
+      "step": 412400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.682356708293308e-05,
+      "loss": 2.5166,
+      "step": 412500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.682036503362152e-05,
+      "loss": 2.487,
+      "step": 412600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6817162984309964e-05,
+      "loss": 2.4917,
+      "step": 412700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.68139609349984e-05,
+      "loss": 2.505,
+      "step": 412800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.6810758885686844e-05,
+      "loss": 2.476,
+      "step": 412900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.680755683637528e-05,
+      "loss": 2.4829,
+      "step": 413000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.4932498931884766,
+      "eval_runtime": 174.2872,
+      "eval_samples_per_second": 57.377,
+      "eval_steps_per_second": 3.586,
+      "step": 413000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.680435478706372e-05,
+      "loss": 2.4614,
+      "step": 413100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.680115273775216e-05,
+      "loss": 2.4694,
+      "step": 413200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.67979506884406e-05,
+      "loss": 2.4825,
+      "step": 413300
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.679474863912904e-05,
+      "loss": 2.4695,
+      "step": 413400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.679154658981748e-05,
+      "loss": 2.499,
+      "step": 413500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.678834454050593e-05,
+      "loss": 2.5021,
+      "step": 413600
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.678514249119437e-05,
+      "loss": 2.5068,
+      "step": 413700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.678194044188281e-05,
+      "loss": 2.4818,
+      "step": 413800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.677873839257125e-05,
+      "loss": 2.4985,
+      "step": 413900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.677553634325969e-05,
+      "loss": 2.4631,
+      "step": 414000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.494476318359375,
+      "eval_runtime": 174.3204,
+      "eval_samples_per_second": 57.366,
+      "eval_steps_per_second": 3.585,
+      "step": 414000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.677233429394813e-05,
+      "loss": 2.4697,
+      "step": 414100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.676913224463657e-05,
+      "loss": 2.4702,
+      "step": 414200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6765930195325013e-05,
+      "loss": 2.492,
+      "step": 414300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6762728146013446e-05,
+      "loss": 2.4991,
+      "step": 414400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.675952609670189e-05,
+      "loss": 2.4868,
+      "step": 414500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.675632404739033e-05,
+      "loss": 2.4767,
+      "step": 414600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.675312199807877e-05,
+      "loss": 2.4825,
+      "step": 414700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.674991994876721e-05,
+      "loss": 2.4673,
+      "step": 414800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.674671789945565e-05,
+      "loss": 2.4693,
+      "step": 414900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.674351585014409e-05,
+      "loss": 2.4845,
+      "step": 415000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.4948008060455322,
+      "eval_runtime": 174.7629,
+      "eval_samples_per_second": 57.22,
+      "eval_steps_per_second": 3.576,
+      "step": 415000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.674031380083253e-05,
+      "loss": 2.4759,
+      "step": 415100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.673711175152098e-05,
+      "loss": 2.5018,
+      "step": 415200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.673390970220942e-05,
+      "loss": 2.4912,
+      "step": 415300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.673070765289786e-05,
+      "loss": 2.4889,
+      "step": 415400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.67275056035863e-05,
+      "loss": 2.4829,
+      "step": 415500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.672430355427474e-05,
+      "loss": 2.4768,
+      "step": 415600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6721101504963177e-05,
+      "loss": 2.4834,
+      "step": 415700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6717899455651616e-05,
+      "loss": 2.4717,
+      "step": 415800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.671469740634006e-05,
+      "loss": 2.4642,
+      "step": 415900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.67114953570285e-05,
+      "loss": 2.5015,
+      "step": 416000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.4942965507507324,
+      "eval_runtime": 173.9314,
+      "eval_samples_per_second": 57.494,
+      "eval_steps_per_second": 3.593,
+      "step": 416000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.670829330771694e-05,
+      "loss": 2.4761,
+      "step": 416100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.670509125840538e-05,
+      "loss": 2.4949,
+      "step": 416200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.670188920909382e-05,
+      "loss": 2.4876,
+      "step": 416300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.669868715978226e-05,
+      "loss": 2.4664,
+      "step": 416400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.66954851104707e-05,
+      "loss": 2.4787,
+      "step": 416500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.669228306115914e-05,
+      "loss": 2.4904,
+      "step": 416600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.668908101184759e-05,
+      "loss": 2.485,
+      "step": 416700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.668587896253603e-05,
+      "loss": 2.4717,
+      "step": 416800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.668267691322447e-05,
+      "loss": 2.4711,
+      "step": 416900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.667947486391291e-05,
+      "loss": 2.5017,
+      "step": 417000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.4940578937530518,
+      "eval_runtime": 177.4234,
+      "eval_samples_per_second": 56.362,
+      "eval_steps_per_second": 3.523,
+      "step": 417000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6676272814601346e-05,
+      "loss": 2.4778,
+      "step": 417100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6673070765289786e-05,
+      "loss": 2.4829,
+      "step": 417200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6669868715978226e-05,
+      "loss": 2.4863,
+      "step": 417300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 2.4672,
+      "step": 417400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.666346461735511e-05,
+      "loss": 2.4826,
+      "step": 417500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.666026256804355e-05,
+      "loss": 2.5087,
+      "step": 417600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.665706051873199e-05,
+      "loss": 2.5071,
+      "step": 417700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.665385846942043e-05,
+      "loss": 2.4953,
+      "step": 417800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.665065642010887e-05,
+      "loss": 2.5056,
+      "step": 417900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.664745437079731e-05,
+      "loss": 2.4985,
+      "step": 418000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.492438554763794,
+      "eval_runtime": 177.1033,
+      "eval_samples_per_second": 56.464,
+      "eval_steps_per_second": 3.529,
+      "step": 418000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.664425232148575e-05,
+      "loss": 2.5036,
+      "step": 418100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.664105027217419e-05,
+      "loss": 2.4859,
+      "step": 418200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.663784822286264e-05,
+      "loss": 2.497,
+      "step": 418300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.663464617355108e-05,
+      "loss": 2.495,
+      "step": 418400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6631444124239516e-05,
+      "loss": 2.4866,
+      "step": 418500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6628242074927956e-05,
+      "loss": 2.4929,
+      "step": 418600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6625040025616396e-05,
+      "loss": 2.5067,
+      "step": 418700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6621837976304836e-05,
+      "loss": 2.4892,
+      "step": 418800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6618635926993275e-05,
+      "loss": 2.4862,
+      "step": 418900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.661543387768172e-05,
+      "loss": 2.4802,
+      "step": 419000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.49405837059021,
+      "eval_runtime": 176.9572,
+      "eval_samples_per_second": 56.511,
+      "eval_steps_per_second": 3.532,
+      "step": 419000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6612231828370155e-05,
+      "loss": 2.4848,
+      "step": 419100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.66090297790586e-05,
+      "loss": 2.4793,
+      "step": 419200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.660582772974704e-05,
+      "loss": 2.4655,
+      "step": 419300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.660262568043548e-05,
+      "loss": 2.478,
+      "step": 419400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.659942363112392e-05,
+      "loss": 2.4943,
+      "step": 419500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.659622158181236e-05,
+      "loss": 2.4792,
+      "step": 419600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.65930195325008e-05,
+      "loss": 2.4713,
+      "step": 419700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.658981748318924e-05,
+      "loss": 2.487,
+      "step": 419800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6586615433877686e-05,
+      "loss": 2.4838,
+      "step": 419900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6583413384566126e-05,
+      "loss": 2.4632,
+      "step": 420000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.495870590209961,
+      "eval_runtime": 177.4486,
+      "eval_samples_per_second": 56.354,
+      "eval_steps_per_second": 3.522,
+      "step": 420000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6580211335254566e-05,
+      "loss": 2.4954,
+      "step": 420100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6577009285943006e-05,
+      "loss": 2.4779,
+      "step": 420200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6573807236631445e-05,
+      "loss": 2.4693,
+      "step": 420300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6570605187319885e-05,
+      "loss": 2.4719,
+      "step": 420400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6567403138008325e-05,
+      "loss": 2.4674,
+      "step": 420500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.656420108869677e-05,
+      "loss": 2.4754,
+      "step": 420600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6560999039385204e-05,
+      "loss": 2.4842,
+      "step": 420700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.655779699007365e-05,
+      "loss": 2.4858,
+      "step": 420800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.655459494076209e-05,
+      "loss": 2.4592,
+      "step": 420900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.655139289145053e-05,
+      "loss": 2.4765,
+      "step": 421000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.4958765506744385,
+      "eval_runtime": 177.6241,
+      "eval_samples_per_second": 56.299,
+      "eval_steps_per_second": 3.519,
+      "step": 421000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.654819084213897e-05,
+      "loss": 2.4862,
+      "step": 421100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.654498879282741e-05,
+      "loss": 2.4507,
+      "step": 421200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6541786743515856e-05,
+      "loss": 2.5029,
+      "step": 421300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.653858469420429e-05,
+      "loss": 2.4713,
+      "step": 421400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6535382644892736e-05,
+      "loss": 2.4763,
+      "step": 421500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6532180595581175e-05,
+      "loss": 2.4569,
+      "step": 421600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6528978546269615e-05,
+      "loss": 2.4754,
+      "step": 421700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6525776496958055e-05,
+      "loss": 2.4744,
+      "step": 421800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6522574447646495e-05,
+      "loss": 2.4856,
+      "step": 421900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6519372398334934e-05,
+      "loss": 2.4698,
+      "step": 422000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.494734287261963,
+      "eval_runtime": 177.2819,
+      "eval_samples_per_second": 56.407,
+      "eval_steps_per_second": 3.525,
+      "step": 422000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6516170349023374e-05,
+      "loss": 2.4684,
+      "step": 422100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.651296829971182e-05,
+      "loss": 2.4747,
+      "step": 422200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6509766250400254e-05,
+      "loss": 2.4901,
+      "step": 422300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.65065642010887e-05,
+      "loss": 2.4759,
+      "step": 422400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.650336215177714e-05,
+      "loss": 2.4759,
+      "step": 422500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.650016010246558e-05,
+      "loss": 2.4831,
+      "step": 422600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.649695805315402e-05,
+      "loss": 2.4787,
+      "step": 422700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.649375600384246e-05,
+      "loss": 2.4861,
+      "step": 422800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6490553954530906e-05,
+      "loss": 2.4596,
+      "step": 422900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.648735190521934e-05,
+      "loss": 2.454,
+      "step": 423000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.49556565284729,
+      "eval_runtime": 177.5164,
+      "eval_samples_per_second": 56.333,
+      "eval_steps_per_second": 3.521,
+      "step": 423000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6484149855907785e-05,
+      "loss": 2.4739,
+      "step": 423100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6480947806596225e-05,
+      "loss": 2.4826,
+      "step": 423200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6477745757284665e-05,
+      "loss": 2.4642,
+      "step": 423300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6474543707973104e-05,
+      "loss": 2.4935,
+      "step": 423400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6471341658661544e-05,
+      "loss": 2.4811,
+      "step": 423500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.646813960934999e-05,
+      "loss": 2.489,
+      "step": 423600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6464937560038423e-05,
+      "loss": 2.4683,
+      "step": 423700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.646173551072687e-05,
+      "loss": 2.4814,
+      "step": 423800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.64585334614153e-05,
+      "loss": 2.4841,
+      "step": 423900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.645533141210375e-05,
+      "loss": 2.4703,
+      "step": 424000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.496413230895996,
+      "eval_runtime": 177.072,
+      "eval_samples_per_second": 56.474,
+      "eval_steps_per_second": 3.53,
+      "step": 424000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.645212936279219e-05,
+      "loss": 2.4733,
+      "step": 424100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.644892731348063e-05,
+      "loss": 2.4829,
+      "step": 424200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.644572526416907e-05,
+      "loss": 2.4814,
+      "step": 424300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.644252321485751e-05,
+      "loss": 2.4794,
+      "step": 424400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6439321165545955e-05,
+      "loss": 2.4672,
+      "step": 424500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.643611911623439e-05,
+      "loss": 2.4783,
+      "step": 424600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6432917066922834e-05,
+      "loss": 2.4736,
+      "step": 424700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6429715017611274e-05,
+      "loss": 2.4833,
+      "step": 424800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6426512968299714e-05,
+      "loss": 2.4814,
+      "step": 424900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6423310918988154e-05,
+      "loss": 2.472,
+      "step": 425000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.4947168827056885,
+      "eval_runtime": 178.8163,
+      "eval_samples_per_second": 55.923,
+      "eval_steps_per_second": 3.495,
+      "step": 425000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.642010886967659e-05,
+      "loss": 2.4883,
+      "step": 425100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.641690682036504e-05,
+      "loss": 2.4865,
+      "step": 425200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.641370477105347e-05,
+      "loss": 2.465,
+      "step": 425300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.641050272174192e-05,
+      "loss": 2.4774,
+      "step": 425400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.640730067243035e-05,
+      "loss": 2.489,
+      "step": 425500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.64040986231188e-05,
+      "loss": 2.4767,
+      "step": 425600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.640089657380724e-05,
+      "loss": 2.4766,
+      "step": 425700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.639769452449568e-05,
+      "loss": 2.4751,
+      "step": 425800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6394492475184125e-05,
+      "loss": 2.4676,
+      "step": 425900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.639129042587256e-05,
+      "loss": 2.5024,
+      "step": 426000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.492905378341675,
+      "eval_runtime": 176.9952,
+      "eval_samples_per_second": 56.499,
+      "eval_steps_per_second": 3.531,
+      "step": 426000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6388088376561004e-05,
+      "loss": 2.4877,
+      "step": 426100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.638488632724944e-05,
+      "loss": 2.4866,
+      "step": 426200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6381684277937884e-05,
+      "loss": 2.462,
+      "step": 426300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6378482228626324e-05,
+      "loss": 2.4671,
+      "step": 426400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.637528017931476e-05,
+      "loss": 2.4859,
+      "step": 426500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.637207813000321e-05,
+      "loss": 2.4792,
+      "step": 426600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.636887608069164e-05,
+      "loss": 2.4829,
+      "step": 426700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.636567403138009e-05,
+      "loss": 2.4976,
+      "step": 426800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.636247198206852e-05,
+      "loss": 2.4798,
+      "step": 426900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.635926993275697e-05,
+      "loss": 2.4688,
+      "step": 427000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.495476722717285,
+      "eval_runtime": 177.9819,
+      "eval_samples_per_second": 56.185,
+      "eval_steps_per_second": 3.512,
+      "step": 427000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.63560678834454e-05,
+      "loss": 2.4749,
+      "step": 427100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.635286583413385e-05,
+      "loss": 2.4752,
+      "step": 427200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.634966378482229e-05,
+      "loss": 2.4922,
+      "step": 427300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.634646173551073e-05,
+      "loss": 2.4857,
+      "step": 427400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6343259686199174e-05,
+      "loss": 2.476,
+      "step": 427500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.634005763688761e-05,
+      "loss": 2.4561,
+      "step": 427600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6336855587576054e-05,
+      "loss": 2.4772,
+      "step": 427700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.633365353826449e-05,
+      "loss": 2.4719,
+      "step": 427800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.633045148895293e-05,
+      "loss": 2.4825,
+      "step": 427900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.632724943964137e-05,
+      "loss": 2.463,
+      "step": 428000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.494075298309326,
+      "eval_runtime": 178.0026,
+      "eval_samples_per_second": 56.179,
+      "eval_steps_per_second": 3.511,
+      "step": 428000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.632404739032981e-05,
+      "loss": 2.459,
+      "step": 428100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.632084534101826e-05,
+      "loss": 2.4766,
+      "step": 428200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.631764329170669e-05,
+      "loss": 2.4788,
+      "step": 428300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.631444124239514e-05,
+      "loss": 2.4721,
+      "step": 428400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.631123919308357e-05,
+      "loss": 2.4855,
+      "step": 428500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.630803714377202e-05,
+      "loss": 2.4738,
+      "step": 428600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.630483509446045e-05,
+      "loss": 2.4772,
+      "step": 428700
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.63016330451489e-05,
+      "loss": 2.467,
+      "step": 428800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.629843099583734e-05,
+      "loss": 2.4587,
+      "step": 428900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.629522894652578e-05,
+      "loss": 2.4758,
+      "step": 429000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.494337320327759,
+      "eval_runtime": 178.7881,
+      "eval_samples_per_second": 55.932,
+      "eval_steps_per_second": 3.496,
+      "step": 429000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6292026897214224e-05,
+      "loss": 2.4677,
+      "step": 429100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6288824847902657e-05,
+      "loss": 2.4757,
+      "step": 429200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.62856227985911e-05,
+      "loss": 2.4604,
+      "step": 429300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.6282420749279536e-05,
+      "loss": 2.4735,
+      "step": 429400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.627921869996798e-05,
+      "loss": 2.4672,
+      "step": 429500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.627601665065642e-05,
+      "loss": 2.4663,
+      "step": 429600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.627281460134486e-05,
+      "loss": 2.4685,
+      "step": 429700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.626961255203331e-05,
+      "loss": 2.4663,
+      "step": 429800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.626641050272174e-05,
+      "loss": 2.4527,
+      "step": 429900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.626320845341019e-05,
+      "loss": 2.4643,
+      "step": 430000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4952518939971924,
+      "eval_runtime": 177.0684,
+      "eval_samples_per_second": 56.475,
+      "eval_steps_per_second": 3.53,
+      "step": 430000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.626000640409862e-05,
+      "loss": 2.4684,
+      "step": 430100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.625680435478707e-05,
+      "loss": 2.4798,
+      "step": 430200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.62536023054755e-05,
+      "loss": 2.4671,
+      "step": 430300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.625040025616395e-05,
+      "loss": 2.4579,
+      "step": 430400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.624719820685239e-05,
+      "loss": 2.4635,
+      "step": 430500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6243996157540826e-05,
+      "loss": 2.4713,
+      "step": 430600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.624079410822927e-05,
+      "loss": 2.4734,
+      "step": 430700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6237592058917706e-05,
+      "loss": 2.476,
+      "step": 430800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.623439000960615e-05,
+      "loss": 2.4674,
+      "step": 430900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6231187960294585e-05,
+      "loss": 2.4588,
+      "step": 431000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4956154823303223,
+      "eval_runtime": 177.2162,
+      "eval_samples_per_second": 56.428,
+      "eval_steps_per_second": 3.527,
+      "step": 431000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.622798591098303e-05,
+      "loss": 2.4642,
+      "step": 431100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.622478386167147e-05,
+      "loss": 2.4684,
+      "step": 431200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.622158181235991e-05,
+      "loss": 2.4802,
+      "step": 431300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.621837976304836e-05,
+      "loss": 2.4736,
+      "step": 431400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.621517771373679e-05,
+      "loss": 2.464,
+      "step": 431500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.621197566442524e-05,
+      "loss": 2.466,
+      "step": 431600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.620877361511367e-05,
+      "loss": 2.4719,
+      "step": 431700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.620557156580212e-05,
+      "loss": 2.4805,
+      "step": 431800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.620236951649055e-05,
+      "loss": 2.4887,
+      "step": 431900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6199167467178996e-05,
+      "loss": 2.4649,
+      "step": 432000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4944067001342773,
+      "eval_runtime": 177.969,
+      "eval_samples_per_second": 56.19,
+      "eval_steps_per_second": 3.512,
+      "step": 432000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6195965417867436e-05,
+      "loss": 2.4812,
+      "step": 432100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6192763368555876e-05,
+      "loss": 2.4749,
+      "step": 432200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.618956131924432e-05,
+      "loss": 2.4669,
+      "step": 432300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6186359269932755e-05,
+      "loss": 2.4679,
+      "step": 432400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.61831572206212e-05,
+      "loss": 2.4874,
+      "step": 432500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6179955171309635e-05,
+      "loss": 2.4755,
+      "step": 432600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.617675312199808e-05,
+      "loss": 2.4708,
+      "step": 432700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.617355107268652e-05,
+      "loss": 2.4744,
+      "step": 432800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.617034902337496e-05,
+      "loss": 2.4797,
+      "step": 432900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.616714697406341e-05,
+      "loss": 2.4551,
+      "step": 433000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.495239734649658,
+      "eval_runtime": 178.15,
+      "eval_samples_per_second": 56.132,
+      "eval_steps_per_second": 3.508,
+      "step": 433000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.616394492475184e-05,
+      "loss": 2.4559,
+      "step": 433100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.616074287544029e-05,
+      "loss": 2.4529,
+      "step": 433200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.615754082612872e-05,
+      "loss": 2.4838,
+      "step": 433300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6154338776817166e-05,
+      "loss": 2.4913,
+      "step": 433400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6151136727505606e-05,
+      "loss": 2.4775,
+      "step": 433500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6147934678194046e-05,
+      "loss": 2.4805,
+      "step": 433600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6144732628882485e-05,
+      "loss": 2.4598,
+      "step": 433700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6141530579570925e-05,
+      "loss": 2.4719,
+      "step": 433800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.613832853025937e-05,
+      "loss": 2.4614,
+      "step": 433900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6135126480947805e-05,
+      "loss": 2.4796,
+      "step": 434000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.493349313735962,
+      "eval_runtime": 178.1065,
+      "eval_samples_per_second": 56.146,
+      "eval_steps_per_second": 3.509,
+      "step": 434000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.613192443163625e-05,
+      "loss": 2.4692,
+      "step": 434100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6128722382324684e-05,
+      "loss": 2.472,
+      "step": 434200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.612552033301313e-05,
+      "loss": 2.4806,
+      "step": 434300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.612231828370157e-05,
+      "loss": 2.5066,
+      "step": 434400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.611911623439001e-05,
+      "loss": 2.4824,
+      "step": 434500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.611591418507846e-05,
+      "loss": 2.4553,
+      "step": 434600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.611271213576689e-05,
+      "loss": 2.4641,
+      "step": 434700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6109510086455336e-05,
+      "loss": 2.4703,
+      "step": 434800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.610630803714377e-05,
+      "loss": 2.4885,
+      "step": 434900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6103105987832216e-05,
+      "loss": 2.4675,
+      "step": 435000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4942822456359863,
+      "eval_runtime": 177.9046,
+      "eval_samples_per_second": 56.21,
+      "eval_steps_per_second": 3.513,
+      "step": 435000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6099903938520655e-05,
+      "loss": 2.4845,
+      "step": 435100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6096701889209095e-05,
+      "loss": 2.4817,
+      "step": 435200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6093499839897535e-05,
+      "loss": 2.4671,
+      "step": 435300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6090297790585975e-05,
+      "loss": 2.4666,
+      "step": 435400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.608709574127442e-05,
+      "loss": 2.4653,
+      "step": 435500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6083893691962854e-05,
+      "loss": 2.4874,
+      "step": 435600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.60806916426513e-05,
+      "loss": 2.4933,
+      "step": 435700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.607748959333974e-05,
+      "loss": 2.4735,
+      "step": 435800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.607428754402818e-05,
+      "loss": 2.4854,
+      "step": 435900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.607108549471662e-05,
+      "loss": 2.4806,
+      "step": 436000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4951186180114746,
+      "eval_runtime": 177.2138,
+      "eval_samples_per_second": 56.429,
+      "eval_steps_per_second": 3.527,
+      "step": 436000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.606788344540506e-05,
+      "loss": 2.4915,
+      "step": 436100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.60646813960935e-05,
+      "loss": 2.4785,
+      "step": 436200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.606147934678194e-05,
+      "loss": 2.4932,
+      "step": 436300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6058277297470385e-05,
+      "loss": 2.4729,
+      "step": 436400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6055075248158825e-05,
+      "loss": 2.5004,
+      "step": 436500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6051873198847265e-05,
+      "loss": 2.4784,
+      "step": 436600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6048671149535705e-05,
+      "loss": 2.4774,
+      "step": 436700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6045469100224144e-05,
+      "loss": 2.4783,
+      "step": 436800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6042267050912584e-05,
+      "loss": 2.4763,
+      "step": 436900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6039065001601024e-05,
+      "loss": 2.4744,
+      "step": 437000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.494119644165039,
+      "eval_runtime": 177.0508,
+      "eval_samples_per_second": 56.481,
+      "eval_steps_per_second": 3.53,
+      "step": 437000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.603586295228947e-05,
+      "loss": 2.4783,
+      "step": 437100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.60326609029779e-05,
+      "loss": 2.4644,
+      "step": 437200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.602945885366635e-05,
+      "loss": 2.478,
+      "step": 437300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.602625680435479e-05,
+      "loss": 2.4572,
+      "step": 437400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.602305475504323e-05,
+      "loss": 2.4817,
+      "step": 437500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.601985270573167e-05,
+      "loss": 2.4597,
+      "step": 437600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.601665065642011e-05,
+      "loss": 2.4773,
+      "step": 437700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.601344860710855e-05,
+      "loss": 2.4608,
+      "step": 437800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.601024655779699e-05,
+      "loss": 2.4653,
+      "step": 437900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6007044508485435e-05,
+      "loss": 2.4651,
+      "step": 438000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4944350719451904,
+      "eval_runtime": 176.4637,
+      "eval_samples_per_second": 56.669,
+      "eval_steps_per_second": 3.542,
+      "step": 438000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6003842459173875e-05,
+      "loss": 2.4738,
+      "step": 438100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6000640409862314e-05,
+      "loss": 2.4739,
+      "step": 438200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5997438360550754e-05,
+      "loss": 2.473,
+      "step": 438300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5994236311239194e-05,
+      "loss": 2.4737,
+      "step": 438400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5991034261927634e-05,
+      "loss": 2.4668,
+      "step": 438500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.598783221261607e-05,
+      "loss": 2.4615,
+      "step": 438600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.598463016330452e-05,
+      "loss": 2.4614,
+      "step": 438700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.598142811399296e-05,
+      "loss": 2.4867,
+      "step": 438800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.59782260646814e-05,
+      "loss": 2.479,
+      "step": 438900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.597502401536984e-05,
+      "loss": 2.4717,
+      "step": 439000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.49422550201416,
+      "eval_runtime": 176.3748,
+      "eval_samples_per_second": 56.697,
+      "eval_steps_per_second": 3.544,
+      "step": 439000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.597182196605828e-05,
+      "loss": 2.4801,
+      "step": 439100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.596861991674672e-05,
+      "loss": 2.4789,
+      "step": 439200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.596541786743516e-05,
+      "loss": 2.4716,
+      "step": 439300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.59622158181236e-05,
+      "loss": 2.4616,
+      "step": 439400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.595901376881204e-05,
+      "loss": 2.4795,
+      "step": 439500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5955811719500484e-05,
+      "loss": 2.482,
+      "step": 439600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5952609670188924e-05,
+      "loss": 2.48,
+      "step": 439700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5949407620877364e-05,
+      "loss": 2.4779,
+      "step": 439800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5946205571565803e-05,
+      "loss": 2.471,
+      "step": 439900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.594300352225424e-05,
+      "loss": 2.4799,
+      "step": 440000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.494257926940918,
+      "eval_runtime": 177.2632,
+      "eval_samples_per_second": 56.413,
+      "eval_steps_per_second": 3.526,
+      "step": 440000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.593980147294268e-05,
+      "loss": 2.4841,
+      "step": 440100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.593659942363112e-05,
+      "loss": 2.4754,
+      "step": 440200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.593339737431957e-05,
+      "loss": 2.4856,
+      "step": 440300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.593019532500801e-05,
+      "loss": 2.4596,
+      "step": 440400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.592699327569645e-05,
+      "loss": 2.4729,
+      "step": 440500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.592379122638489e-05,
+      "loss": 2.4772,
+      "step": 440600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.592058917707333e-05,
+      "loss": 2.4668,
+      "step": 440700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.591738712776177e-05,
+      "loss": 2.4635,
+      "step": 440800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.591418507845021e-05,
+      "loss": 2.481,
+      "step": 440900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.591098302913865e-05,
+      "loss": 2.4808,
+      "step": 441000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4937944412231445,
+      "eval_runtime": 176.7561,
+      "eval_samples_per_second": 56.575,
+      "eval_steps_per_second": 3.536,
+      "step": 441000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5907780979827094e-05,
+      "loss": 2.4661,
+      "step": 441100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5904578930515534e-05,
+      "loss": 2.4612,
+      "step": 441200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.590137688120397e-05,
+      "loss": 2.4688,
+      "step": 441300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.589817483189241e-05,
+      "loss": 2.4639,
+      "step": 441400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.589497278258085e-05,
+      "loss": 2.4823,
+      "step": 441500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.589177073326929e-05,
+      "loss": 2.464,
+      "step": 441600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.588856868395773e-05,
+      "loss": 2.4781,
+      "step": 441700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.588536663464617e-05,
+      "loss": 2.476,
+      "step": 441800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.588216458533462e-05,
+      "loss": 2.4694,
+      "step": 441900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.587896253602306e-05,
+      "loss": 2.4791,
+      "step": 442000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4932377338409424,
+      "eval_runtime": 177.3647,
+      "eval_samples_per_second": 56.381,
+      "eval_steps_per_second": 3.524,
+      "step": 442000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.58757604867115e-05,
+      "loss": 2.4855,
+      "step": 442100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.587255843739994e-05,
+      "loss": 2.4666,
+      "step": 442200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.586935638808838e-05,
+      "loss": 2.4782,
+      "step": 442300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.586615433877682e-05,
+      "loss": 2.4878,
+      "step": 442400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.586295228946526e-05,
+      "loss": 2.4789,
+      "step": 442500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.58597502401537e-05,
+      "loss": 2.4673,
+      "step": 442600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.585654819084214e-05,
+      "loss": 2.4407,
+      "step": 442700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.585334614153058e-05,
+      "loss": 2.4581,
+      "step": 442800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.585014409221902e-05,
+      "loss": 2.4714,
+      "step": 442900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.584694204290746e-05,
+      "loss": 2.4655,
+      "step": 443000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4938127994537354,
+      "eval_runtime": 178.5522,
+      "eval_samples_per_second": 56.006,
+      "eval_steps_per_second": 3.5,
+      "step": 443000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.58437399935959e-05,
+      "loss": 2.4742,
+      "step": 443100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.584053794428434e-05,
+      "loss": 2.4513,
+      "step": 443200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.583733589497278e-05,
+      "loss": 2.4779,
+      "step": 443300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.583413384566123e-05,
+      "loss": 2.4659,
+      "step": 443400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.583093179634967e-05,
+      "loss": 2.4824,
+      "step": 443500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.582772974703811e-05,
+      "loss": 2.4671,
+      "step": 443600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.582452769772655e-05,
+      "loss": 2.4798,
+      "step": 443700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.582132564841499e-05,
+      "loss": 2.4541,
+      "step": 443800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.581812359910343e-05,
+      "loss": 2.4834,
+      "step": 443900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5814921549791867e-05,
+      "loss": 2.468,
+      "step": 444000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.492299795150757,
+      "eval_runtime": 178.808,
+      "eval_samples_per_second": 55.926,
+      "eval_steps_per_second": 3.495,
+      "step": 444000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.581171950048031e-05,
+      "loss": 2.4625,
+      "step": 444100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.5808517451168746e-05,
+      "loss": 2.4731,
+      "step": 444200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.580531540185719e-05,
+      "loss": 2.4681,
+      "step": 444300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.580211335254563e-05,
+      "loss": 2.4664,
+      "step": 444400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.579891130323407e-05,
+      "loss": 2.4502,
+      "step": 444500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.579570925392251e-05,
+      "loss": 2.4593,
+      "step": 444600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.579250720461095e-05,
+      "loss": 2.4739,
+      "step": 444700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.578930515529939e-05,
+      "loss": 2.4491,
+      "step": 444800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.578610310598783e-05,
+      "loss": 2.4718,
+      "step": 444900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.578290105667628e-05,
+      "loss": 2.4559,
+      "step": 445000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.4955239295959473,
+      "eval_runtime": 177.5848,
+      "eval_samples_per_second": 56.311,
+      "eval_steps_per_second": 3.519,
+      "step": 445000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.577969900736472e-05,
+      "loss": 2.4634,
+      "step": 445100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.577649695805316e-05,
+      "loss": 2.4743,
+      "step": 445200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.57732949087416e-05,
+      "loss": 2.4673,
+      "step": 445300
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5770092859430036e-05,
+      "loss": 2.469,
+      "step": 445400
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5766890810118476e-05,
+      "loss": 2.4642,
+      "step": 445500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5763688760806916e-05,
+      "loss": 2.4439,
+      "step": 445600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.576048671149536e-05,
+      "loss": 2.4756,
+      "step": 445700
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5757284662183795e-05,
+      "loss": 2.4627,
+      "step": 445800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.575408261287224e-05,
+      "loss": 2.457,
+      "step": 445900
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.575088056356068e-05,
+      "loss": 2.4831,
+      "step": 446000
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 2.49298095703125,
+      "eval_runtime": 176.8362,
+      "eval_samples_per_second": 56.55,
+      "eval_steps_per_second": 3.534,
+      "step": 446000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.574767851424912e-05,
+      "loss": 2.4626,
+      "step": 446100
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.574447646493756e-05,
+      "loss": 2.4478,
+      "step": 446200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5741274415626e-05,
+      "loss": 2.4569,
+      "step": 446300
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.573807236631445e-05,
+      "loss": 2.4704,
+      "step": 446400
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.573487031700288e-05,
+      "loss": 2.4993,
+      "step": 446500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.573166826769133e-05,
+      "loss": 2.4508,
+      "step": 446600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.572846621837977e-05,
+      "loss": 2.4603,
+      "step": 446700
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5725264169068206e-05,
+      "loss": 2.4818,
+      "step": 446800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5722062119756646e-05,
+      "loss": 2.4838,
+      "step": 446900
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5718860070445086e-05,
+      "loss": 2.4736,
+      "step": 447000
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 2.4949324131011963,
+      "eval_runtime": 177.2099,
+      "eval_samples_per_second": 56.43,
+      "eval_steps_per_second": 3.527,
+      "step": 447000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5715658021133526e-05,
+      "loss": 2.4481,
+      "step": 447100
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5712455971821965e-05,
+      "loss": 2.4844,
+      "step": 447200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.570925392251041e-05,
+      "loss": 2.4795,
+      "step": 447300
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5706051873198845e-05,
+      "loss": 2.46,
+      "step": 447400
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.570284982388729e-05,
+      "loss": 2.4569,
+      "step": 447500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.569964777457573e-05,
+      "loss": 2.4753,
+      "step": 447600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.569644572526417e-05,
+      "loss": 2.4581,
+      "step": 447700
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.569324367595261e-05,
+      "loss": 2.4456,
+      "step": 447800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.569004162664105e-05,
+      "loss": 2.4647,
+      "step": 447900
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.56868395773295e-05,
+      "loss": 2.4473,
+      "step": 448000
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 2.496084690093994,
+      "eval_runtime": 173.9796,
+      "eval_samples_per_second": 57.478,
+      "eval_steps_per_second": 3.592,
+      "step": 448000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.568363752801793e-05,
+      "loss": 2.4676,
+      "step": 448100
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5680435478706376e-05,
+      "loss": 2.4709,
+      "step": 448200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5677233429394816e-05,
+      "loss": 2.4867,
+      "step": 448300
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5674031380083256e-05,
+      "loss": 2.4506,
+      "step": 448400
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5670829330771695e-05,
+      "loss": 2.4687,
+      "step": 448500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5667627281460135e-05,
+      "loss": 2.4423,
+      "step": 448600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.566442523214858e-05,
+      "loss": 2.4716,
+      "step": 448700
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5661223182837015e-05,
+      "loss": 2.4638,
+      "step": 448800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.565802113352546e-05,
+      "loss": 2.4549,
+      "step": 448900
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.5654819084213894e-05,
+      "loss": 2.4655,
+      "step": 449000
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 2.494385242462158,
+      "eval_runtime": 173.7929,
+      "eval_samples_per_second": 57.54,
+      "eval_steps_per_second": 3.596,
+      "step": 449000
+    }
+  ],
+  "max_steps": 1562500,
+  "num_train_epochs": 1,
+  "total_flos": 1.229824204996608e+19,
+  "trial_name": null,
+  "trial_params": null
+}