{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "global_step": 180294, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.986133759304248e-05, "loss": 2.7631, "step": 500 }, { "epoch": 0.02, "learning_rate": 4.9722675186084955e-05, "loss": 1.9875, "step": 1000 }, { "epoch": 0.02, "learning_rate": 4.9584012779127425e-05, "loss": 1.7721, "step": 1500 }, { "epoch": 0.03, "learning_rate": 4.94453503721699e-05, "loss": 1.6142, "step": 2000 }, { "epoch": 0.04, "learning_rate": 4.930668796521238e-05, "loss": 1.5159, "step": 2500 }, { "epoch": 0.05, "learning_rate": 4.9168025558254854e-05, "loss": 1.4552, "step": 3000 }, { "epoch": 0.06, "learning_rate": 4.902936315129733e-05, "loss": 1.4205, "step": 3500 }, { "epoch": 0.07, "learning_rate": 4.88907007443398e-05, "loss": 1.355, "step": 4000 }, { "epoch": 0.07, "learning_rate": 4.875203833738228e-05, "loss": 1.3154, "step": 4500 }, { "epoch": 0.08, "learning_rate": 4.861337593042475e-05, "loss": 1.2864, "step": 5000 }, { "epoch": 0.09, "learning_rate": 4.847471352346723e-05, "loss": 1.2668, "step": 5500 }, { "epoch": 0.1, "learning_rate": 4.83360511165097e-05, "loss": 1.2525, "step": 6000 }, { "epoch": 0.11, "learning_rate": 4.8197388709552176e-05, "loss": 1.2474, "step": 6500 }, { "epoch": 0.12, "learning_rate": 4.805872630259465e-05, "loss": 1.2006, "step": 7000 }, { "epoch": 0.12, "learning_rate": 4.792006389563713e-05, "loss": 1.1849, "step": 7500 }, { "epoch": 0.13, "learning_rate": 4.7781401488679605e-05, "loss": 1.1786, "step": 8000 }, { "epoch": 0.14, "learning_rate": 4.7642739081722075e-05, "loss": 1.1462, "step": 8500 }, { "epoch": 0.15, "learning_rate": 4.750407667476455e-05, "loss": 1.1314, "step": 9000 }, { "epoch": 0.16, "learning_rate": 4.736541426780703e-05, "loss": 1.1171, "step": 9500 }, { "epoch": 0.17, "learning_rate": 4.7226751860849504e-05, "loss": 1.1162, "step": 10000 }, { "epoch": 0.17, "learning_rate": 4.708808945389198e-05, "loss": 1.1141, "step": 10500 }, { "epoch": 0.18, "learning_rate": 4.694942704693445e-05, "loss": 1.081, "step": 11000 }, { "epoch": 0.19, "learning_rate": 4.6810764639976926e-05, "loss": 1.0808, "step": 11500 }, { "epoch": 0.2, "learning_rate": 4.66721022330194e-05, "loss": 1.078, "step": 12000 }, { "epoch": 0.21, "learning_rate": 4.653343982606188e-05, "loss": 1.0797, "step": 12500 }, { "epoch": 0.22, "learning_rate": 4.639477741910435e-05, "loss": 1.0495, "step": 13000 }, { "epoch": 0.22, "learning_rate": 4.6256115012146825e-05, "loss": 1.0576, "step": 13500 }, { "epoch": 0.23, "learning_rate": 4.61174526051893e-05, "loss": 1.0461, "step": 14000 }, { "epoch": 0.24, "learning_rate": 4.597879019823178e-05, "loss": 1.0314, "step": 14500 }, { "epoch": 0.25, "learning_rate": 4.5840127791274255e-05, "loss": 1.0223, "step": 15000 }, { "epoch": 0.26, "learning_rate": 4.5701465384316724e-05, "loss": 1.016, "step": 15500 }, { "epoch": 0.27, "learning_rate": 4.55628029773592e-05, "loss": 1.009, "step": 16000 }, { "epoch": 0.27, "learning_rate": 4.542414057040168e-05, "loss": 1.0033, "step": 16500 }, { "epoch": 0.28, "learning_rate": 4.5285478163444154e-05, "loss": 0.9954, "step": 17000 }, { "epoch": 0.29, "learning_rate": 4.514681575648663e-05, "loss": 0.9986, "step": 17500 }, { "epoch": 0.3, "learning_rate": 4.50081533495291e-05, "loss": 0.9932, "step": 18000 }, { "epoch": 0.31, "learning_rate": 4.4869490942571576e-05, "loss": 0.9877, "step": 18500 }, { "epoch": 0.32, "learning_rate": 4.473082853561405e-05, "loss": 0.9707, "step": 19000 }, { "epoch": 0.32, "learning_rate": 4.459216612865653e-05, "loss": 0.9758, "step": 19500 }, { "epoch": 0.33, "learning_rate": 4.4453503721699005e-05, "loss": 0.9739, "step": 20000 }, { "epoch": 0.34, "learning_rate": 4.431484131474148e-05, "loss": 0.9785, "step": 20500 }, { "epoch": 0.35, "learning_rate": 4.417617890778396e-05, "loss": 0.9539, "step": 21000 }, { "epoch": 0.36, "learning_rate": 4.403751650082643e-05, "loss": 0.9442, "step": 21500 }, { "epoch": 0.37, "learning_rate": 4.3898854093868904e-05, "loss": 0.9418, "step": 22000 }, { "epoch": 0.37, "learning_rate": 4.376019168691138e-05, "loss": 0.9513, "step": 22500 }, { "epoch": 0.38, "learning_rate": 4.362152927995386e-05, "loss": 0.9401, "step": 23000 }, { "epoch": 0.39, "learning_rate": 4.3482866872996334e-05, "loss": 0.9386, "step": 23500 }, { "epoch": 0.4, "learning_rate": 4.334420446603881e-05, "loss": 0.9331, "step": 24000 }, { "epoch": 0.41, "learning_rate": 4.3205542059081287e-05, "loss": 0.925, "step": 24500 }, { "epoch": 0.42, "learning_rate": 4.3066879652123756e-05, "loss": 0.9387, "step": 25000 }, { "epoch": 0.42, "learning_rate": 4.292821724516623e-05, "loss": 0.927, "step": 25500 }, { "epoch": 0.43, "learning_rate": 4.278955483820871e-05, "loss": 0.923, "step": 26000 }, { "epoch": 0.44, "learning_rate": 4.2650892431251185e-05, "loss": 0.9103, "step": 26500 }, { "epoch": 0.45, "learning_rate": 4.2512230024293655e-05, "loss": 0.9135, "step": 27000 }, { "epoch": 0.46, "learning_rate": 4.237356761733613e-05, "loss": 0.9107, "step": 27500 }, { "epoch": 0.47, "learning_rate": 4.223490521037861e-05, "loss": 0.8941, "step": 28000 }, { "epoch": 0.47, "learning_rate": 4.2096242803421084e-05, "loss": 0.9239, "step": 28500 }, { "epoch": 0.48, "learning_rate": 4.195758039646356e-05, "loss": 0.9114, "step": 29000 }, { "epoch": 0.49, "learning_rate": 4.181891798950603e-05, "loss": 0.9004, "step": 29500 }, { "epoch": 0.5, "learning_rate": 4.168025558254851e-05, "loss": 0.9063, "step": 30000 }, { "epoch": 0.51, "learning_rate": 4.154159317559098e-05, "loss": 0.8853, "step": 30500 }, { "epoch": 0.52, "learning_rate": 4.140293076863346e-05, "loss": 0.8797, "step": 31000 }, { "epoch": 0.52, "learning_rate": 4.1264268361675936e-05, "loss": 0.8929, "step": 31500 }, { "epoch": 0.53, "learning_rate": 4.1125605954718406e-05, "loss": 0.8907, "step": 32000 }, { "epoch": 0.54, "learning_rate": 4.098694354776088e-05, "loss": 0.8739, "step": 32500 }, { "epoch": 0.55, "learning_rate": 4.084828114080336e-05, "loss": 0.8813, "step": 33000 }, { "epoch": 0.56, "learning_rate": 4.0709618733845835e-05, "loss": 0.8756, "step": 33500 }, { "epoch": 0.57, "learning_rate": 4.0570956326888305e-05, "loss": 0.8758, "step": 34000 }, { "epoch": 0.57, "learning_rate": 4.043229391993078e-05, "loss": 0.8614, "step": 34500 }, { "epoch": 0.58, "learning_rate": 4.029363151297326e-05, "loss": 0.8682, "step": 35000 }, { "epoch": 0.59, "learning_rate": 4.0154969106015734e-05, "loss": 0.8782, "step": 35500 }, { "epoch": 0.6, "learning_rate": 4.001630669905821e-05, "loss": 0.8633, "step": 36000 }, { "epoch": 0.61, "learning_rate": 3.987764429210068e-05, "loss": 0.8728, "step": 36500 }, { "epoch": 0.62, "learning_rate": 3.973898188514316e-05, "loss": 0.8588, "step": 37000 }, { "epoch": 0.62, "learning_rate": 3.960031947818563e-05, "loss": 0.8656, "step": 37500 }, { "epoch": 0.63, "learning_rate": 3.946165707122811e-05, "loss": 0.8619, "step": 38000 }, { "epoch": 0.64, "learning_rate": 3.9322994664270586e-05, "loss": 0.8472, "step": 38500 }, { "epoch": 0.65, "learning_rate": 3.9184332257313056e-05, "loss": 0.8568, "step": 39000 }, { "epoch": 0.66, "learning_rate": 3.904566985035553e-05, "loss": 0.8561, "step": 39500 }, { "epoch": 0.67, "learning_rate": 3.890700744339801e-05, "loss": 0.8519, "step": 40000 }, { "epoch": 0.67, "learning_rate": 3.8768345036440485e-05, "loss": 0.8442, "step": 40500 }, { "epoch": 0.68, "learning_rate": 3.8629682629482955e-05, "loss": 0.8522, "step": 41000 }, { "epoch": 0.69, "learning_rate": 3.849102022252543e-05, "loss": 0.8414, "step": 41500 }, { "epoch": 0.7, "learning_rate": 3.835235781556791e-05, "loss": 0.8303, "step": 42000 }, { "epoch": 0.71, "learning_rate": 3.8213695408610384e-05, "loss": 0.841, "step": 42500 }, { "epoch": 0.72, "learning_rate": 3.807503300165286e-05, "loss": 0.8283, "step": 43000 }, { "epoch": 0.72, "learning_rate": 3.793637059469533e-05, "loss": 0.8377, "step": 43500 }, { "epoch": 0.73, "learning_rate": 3.7797708187737806e-05, "loss": 0.8257, "step": 44000 }, { "epoch": 0.74, "learning_rate": 3.765904578078028e-05, "loss": 0.8328, "step": 44500 }, { "epoch": 0.75, "learning_rate": 3.752038337382276e-05, "loss": 0.8217, "step": 45000 }, { "epoch": 0.76, "learning_rate": 3.7381720966865236e-05, "loss": 0.8237, "step": 45500 }, { "epoch": 0.77, "learning_rate": 3.7243058559907705e-05, "loss": 0.8186, "step": 46000 }, { "epoch": 0.77, "learning_rate": 3.710439615295018e-05, "loss": 0.8302, "step": 46500 }, { "epoch": 0.78, "learning_rate": 3.696573374599266e-05, "loss": 0.8221, "step": 47000 }, { "epoch": 0.79, "learning_rate": 3.6827071339035135e-05, "loss": 0.8221, "step": 47500 }, { "epoch": 0.8, "learning_rate": 3.6688408932077604e-05, "loss": 0.8258, "step": 48000 }, { "epoch": 0.81, "learning_rate": 3.654974652512008e-05, "loss": 0.8289, "step": 48500 }, { "epoch": 0.82, "learning_rate": 3.641108411816256e-05, "loss": 0.8089, "step": 49000 }, { "epoch": 0.82, "learning_rate": 3.6272421711205034e-05, "loss": 0.8113, "step": 49500 }, { "epoch": 0.83, "learning_rate": 3.613375930424751e-05, "loss": 0.8075, "step": 50000 }, { "epoch": 0.84, "learning_rate": 3.599509689728998e-05, "loss": 0.8048, "step": 50500 }, { "epoch": 0.85, "learning_rate": 3.5856434490332456e-05, "loss": 0.8089, "step": 51000 }, { "epoch": 0.86, "learning_rate": 3.571777208337493e-05, "loss": 0.8097, "step": 51500 }, { "epoch": 0.87, "learning_rate": 3.557910967641741e-05, "loss": 0.7908, "step": 52000 }, { "epoch": 0.87, "learning_rate": 3.5440447269459885e-05, "loss": 0.8128, "step": 52500 }, { "epoch": 0.88, "learning_rate": 3.5301784862502355e-05, "loss": 0.8109, "step": 53000 }, { "epoch": 0.89, "learning_rate": 3.516312245554483e-05, "loss": 0.804, "step": 53500 }, { "epoch": 0.9, "learning_rate": 3.502446004858731e-05, "loss": 0.805, "step": 54000 }, { "epoch": 0.91, "learning_rate": 3.4885797641629784e-05, "loss": 0.7859, "step": 54500 }, { "epoch": 0.92, "learning_rate": 3.4747135234672254e-05, "loss": 0.7881, "step": 55000 }, { "epoch": 0.92, "learning_rate": 3.460847282771473e-05, "loss": 0.8019, "step": 55500 }, { "epoch": 0.93, "learning_rate": 3.446981042075721e-05, "loss": 0.7831, "step": 56000 }, { "epoch": 0.94, "learning_rate": 3.433114801379968e-05, "loss": 0.7946, "step": 56500 }, { "epoch": 0.95, "learning_rate": 3.419248560684216e-05, "loss": 0.7919, "step": 57000 }, { "epoch": 0.96, "learning_rate": 3.405382319988463e-05, "loss": 0.7894, "step": 57500 }, { "epoch": 0.97, "learning_rate": 3.3915160792927106e-05, "loss": 0.7884, "step": 58000 }, { "epoch": 0.97, "learning_rate": 3.377649838596958e-05, "loss": 0.7873, "step": 58500 }, { "epoch": 0.98, "learning_rate": 3.363783597901206e-05, "loss": 0.799, "step": 59000 }, { "epoch": 0.99, "learning_rate": 3.3499173572054535e-05, "loss": 0.7809, "step": 59500 }, { "epoch": 1.0, "learning_rate": 3.336051116509701e-05, "loss": 0.7821, "step": 60000 }, { "epoch": 1.01, "learning_rate": 3.322184875813948e-05, "loss": 0.7348, "step": 60500 }, { "epoch": 1.02, "learning_rate": 3.308318635118196e-05, "loss": 0.7154, "step": 61000 }, { "epoch": 1.02, "learning_rate": 3.2944523944224434e-05, "loss": 0.7056, "step": 61500 }, { "epoch": 1.03, "learning_rate": 3.280586153726691e-05, "loss": 0.701, "step": 62000 }, { "epoch": 1.04, "learning_rate": 3.266719913030939e-05, "loss": 0.7071, "step": 62500 }, { "epoch": 1.05, "learning_rate": 3.252853672335186e-05, "loss": 0.7016, "step": 63000 }, { "epoch": 1.06, "learning_rate": 3.238987431639434e-05, "loss": 0.7186, "step": 63500 }, { "epoch": 1.06, "learning_rate": 3.225121190943681e-05, "loss": 0.7182, "step": 64000 }, { "epoch": 1.07, "learning_rate": 3.2112549502479286e-05, "loss": 0.7152, "step": 64500 }, { "epoch": 1.08, "learning_rate": 3.197388709552176e-05, "loss": 0.7057, "step": 65000 }, { "epoch": 1.09, "learning_rate": 3.183522468856424e-05, "loss": 0.6946, "step": 65500 }, { "epoch": 1.1, "learning_rate": 3.1696562281606715e-05, "loss": 0.6964, "step": 66000 }, { "epoch": 1.11, "learning_rate": 3.155789987464919e-05, "loss": 0.7067, "step": 66500 }, { "epoch": 1.11, "learning_rate": 3.141923746769166e-05, "loss": 0.7076, "step": 67000 }, { "epoch": 1.12, "learning_rate": 3.128057506073414e-05, "loss": 0.7178, "step": 67500 }, { "epoch": 1.13, "learning_rate": 3.1141912653776614e-05, "loss": 0.6996, "step": 68000 }, { "epoch": 1.14, "learning_rate": 3.100325024681909e-05, "loss": 0.7006, "step": 68500 }, { "epoch": 1.15, "learning_rate": 3.086458783986156e-05, "loss": 0.69, "step": 69000 }, { "epoch": 1.16, "learning_rate": 3.0725925432904037e-05, "loss": 0.7071, "step": 69500 }, { "epoch": 1.16, "learning_rate": 3.058726302594651e-05, "loss": 0.7014, "step": 70000 }, { "epoch": 1.17, "learning_rate": 3.0448600618988986e-05, "loss": 0.7051, "step": 70500 }, { "epoch": 1.18, "learning_rate": 3.0309938212031462e-05, "loss": 0.7074, "step": 71000 }, { "epoch": 1.19, "learning_rate": 3.0171275805073935e-05, "loss": 0.7072, "step": 71500 }, { "epoch": 1.2, "learning_rate": 3.0032613398116412e-05, "loss": 0.7031, "step": 72000 }, { "epoch": 1.21, "learning_rate": 2.989395099115889e-05, "loss": 0.7106, "step": 72500 }, { "epoch": 1.21, "learning_rate": 2.975528858420136e-05, "loss": 0.7008, "step": 73000 }, { "epoch": 1.22, "learning_rate": 2.9616626177243838e-05, "loss": 0.7066, "step": 73500 }, { "epoch": 1.23, "learning_rate": 2.947796377028631e-05, "loss": 0.7051, "step": 74000 }, { "epoch": 1.24, "learning_rate": 2.9339301363328787e-05, "loss": 0.7114, "step": 74500 }, { "epoch": 1.25, "learning_rate": 2.9200638956371264e-05, "loss": 0.6977, "step": 75000 }, { "epoch": 1.26, "learning_rate": 2.906197654941374e-05, "loss": 0.6933, "step": 75500 }, { "epoch": 1.26, "learning_rate": 2.892331414245621e-05, "loss": 0.6886, "step": 76000 }, { "epoch": 1.27, "learning_rate": 2.8784651735498686e-05, "loss": 0.7023, "step": 76500 }, { "epoch": 1.28, "learning_rate": 2.8645989328541163e-05, "loss": 0.705, "step": 77000 }, { "epoch": 1.29, "learning_rate": 2.850732692158364e-05, "loss": 0.6944, "step": 77500 }, { "epoch": 1.3, "learning_rate": 2.8368664514626116e-05, "loss": 0.7034, "step": 78000 }, { "epoch": 1.31, "learning_rate": 2.8230002107668585e-05, "loss": 0.6887, "step": 78500 }, { "epoch": 1.31, "learning_rate": 2.809133970071106e-05, "loss": 0.7061, "step": 79000 }, { "epoch": 1.32, "learning_rate": 2.7952677293753538e-05, "loss": 0.6973, "step": 79500 }, { "epoch": 1.33, "learning_rate": 2.7814014886796014e-05, "loss": 0.6818, "step": 80000 }, { "epoch": 1.34, "learning_rate": 2.767535247983849e-05, "loss": 0.6778, "step": 80500 }, { "epoch": 1.35, "learning_rate": 2.753669007288096e-05, "loss": 0.6856, "step": 81000 }, { "epoch": 1.36, "learning_rate": 2.7398027665923437e-05, "loss": 0.6878, "step": 81500 }, { "epoch": 1.36, "learning_rate": 2.7259365258965913e-05, "loss": 0.6899, "step": 82000 }, { "epoch": 1.37, "learning_rate": 2.712070285200839e-05, "loss": 0.6949, "step": 82500 }, { "epoch": 1.38, "learning_rate": 2.698204044505086e-05, "loss": 0.6895, "step": 83000 }, { "epoch": 1.39, "learning_rate": 2.6843378038093336e-05, "loss": 0.6728, "step": 83500 }, { "epoch": 1.4, "learning_rate": 2.6704715631135812e-05, "loss": 0.6826, "step": 84000 }, { "epoch": 1.41, "learning_rate": 2.656605322417829e-05, "loss": 0.6925, "step": 84500 }, { "epoch": 1.41, "learning_rate": 2.6427390817220765e-05, "loss": 0.6953, "step": 85000 }, { "epoch": 1.42, "learning_rate": 2.6288728410263235e-05, "loss": 0.6911, "step": 85500 }, { "epoch": 1.43, "learning_rate": 2.615006600330571e-05, "loss": 0.6857, "step": 86000 }, { "epoch": 1.44, "learning_rate": 2.6011403596348188e-05, "loss": 0.6787, "step": 86500 }, { "epoch": 1.45, "learning_rate": 2.5872741189390664e-05, "loss": 0.6742, "step": 87000 }, { "epoch": 1.46, "learning_rate": 2.573407878243314e-05, "loss": 0.6711, "step": 87500 }, { "epoch": 1.46, "learning_rate": 2.559541637547561e-05, "loss": 0.6719, "step": 88000 }, { "epoch": 1.47, "learning_rate": 2.5456753968518087e-05, "loss": 0.6808, "step": 88500 }, { "epoch": 1.48, "learning_rate": 2.5318091561560563e-05, "loss": 0.6815, "step": 89000 }, { "epoch": 1.49, "learning_rate": 2.517942915460304e-05, "loss": 0.6819, "step": 89500 }, { "epoch": 1.5, "learning_rate": 2.5040766747645513e-05, "loss": 0.6823, "step": 90000 }, { "epoch": 1.51, "learning_rate": 2.4902104340687986e-05, "loss": 0.6833, "step": 90500 }, { "epoch": 1.51, "learning_rate": 2.4763441933730462e-05, "loss": 0.6659, "step": 91000 }, { "epoch": 1.52, "learning_rate": 2.462477952677294e-05, "loss": 0.6755, "step": 91500 }, { "epoch": 1.53, "learning_rate": 2.4486117119815415e-05, "loss": 0.6567, "step": 92000 }, { "epoch": 1.54, "learning_rate": 2.4347454712857888e-05, "loss": 0.6739, "step": 92500 }, { "epoch": 1.55, "learning_rate": 2.4208792305900364e-05, "loss": 0.6695, "step": 93000 }, { "epoch": 1.56, "learning_rate": 2.407012989894284e-05, "loss": 0.6727, "step": 93500 }, { "epoch": 1.56, "learning_rate": 2.3931467491985314e-05, "loss": 0.6748, "step": 94000 }, { "epoch": 1.57, "learning_rate": 2.379280508502779e-05, "loss": 0.689, "step": 94500 }, { "epoch": 1.58, "learning_rate": 2.3654142678070267e-05, "loss": 0.6685, "step": 95000 }, { "epoch": 1.59, "learning_rate": 2.351548027111274e-05, "loss": 0.6692, "step": 95500 }, { "epoch": 1.6, "learning_rate": 2.3376817864155216e-05, "loss": 0.6645, "step": 96000 }, { "epoch": 1.61, "learning_rate": 2.323815545719769e-05, "loss": 0.6567, "step": 96500 }, { "epoch": 1.61, "learning_rate": 2.3099493050240166e-05, "loss": 0.6744, "step": 97000 }, { "epoch": 1.62, "learning_rate": 2.296083064328264e-05, "loss": 0.6607, "step": 97500 }, { "epoch": 1.63, "learning_rate": 2.2822168236325115e-05, "loss": 0.6663, "step": 98000 }, { "epoch": 1.64, "learning_rate": 2.268350582936759e-05, "loss": 0.6625, "step": 98500 }, { "epoch": 1.65, "learning_rate": 2.2544843422410065e-05, "loss": 0.6608, "step": 99000 }, { "epoch": 1.66, "learning_rate": 2.240618101545254e-05, "loss": 0.6746, "step": 99500 }, { "epoch": 1.66, "learning_rate": 2.2267518608495014e-05, "loss": 0.653, "step": 100000 }, { "epoch": 1.67, "learning_rate": 2.212885620153749e-05, "loss": 0.6539, "step": 100500 }, { "epoch": 1.68, "learning_rate": 2.1990193794579964e-05, "loss": 0.6706, "step": 101000 }, { "epoch": 1.69, "learning_rate": 2.185153138762244e-05, "loss": 0.6619, "step": 101500 }, { "epoch": 1.7, "learning_rate": 2.1712868980664916e-05, "loss": 0.6659, "step": 102000 }, { "epoch": 1.71, "learning_rate": 2.157420657370739e-05, "loss": 0.6628, "step": 102500 }, { "epoch": 1.71, "learning_rate": 2.1435544166749866e-05, "loss": 0.6568, "step": 103000 }, { "epoch": 1.72, "learning_rate": 2.129688175979234e-05, "loss": 0.6644, "step": 103500 }, { "epoch": 1.73, "learning_rate": 2.1158219352834815e-05, "loss": 0.655, "step": 104000 }, { "epoch": 1.74, "learning_rate": 2.101955694587729e-05, "loss": 0.6536, "step": 104500 }, { "epoch": 1.75, "learning_rate": 2.0880894538919765e-05, "loss": 0.6755, "step": 105000 }, { "epoch": 1.76, "learning_rate": 2.074223213196224e-05, "loss": 0.6575, "step": 105500 }, { "epoch": 1.76, "learning_rate": 2.0603569725004714e-05, "loss": 0.6535, "step": 106000 }, { "epoch": 1.77, "learning_rate": 2.046490731804719e-05, "loss": 0.6531, "step": 106500 }, { "epoch": 1.78, "learning_rate": 2.0326244911089664e-05, "loss": 0.656, "step": 107000 }, { "epoch": 1.79, "learning_rate": 2.018758250413214e-05, "loss": 0.6459, "step": 107500 }, { "epoch": 1.8, "learning_rate": 2.0048920097174613e-05, "loss": 0.6646, "step": 108000 }, { "epoch": 1.81, "learning_rate": 1.991025769021709e-05, "loss": 0.6518, "step": 108500 }, { "epoch": 1.81, "learning_rate": 1.9771595283259566e-05, "loss": 0.6539, "step": 109000 }, { "epoch": 1.82, "learning_rate": 1.963293287630204e-05, "loss": 0.6568, "step": 109500 }, { "epoch": 1.83, "learning_rate": 1.9494270469344516e-05, "loss": 0.6647, "step": 110000 }, { "epoch": 1.84, "learning_rate": 1.935560806238699e-05, "loss": 0.6682, "step": 110500 }, { "epoch": 1.85, "learning_rate": 1.9216945655429465e-05, "loss": 0.6564, "step": 111000 }, { "epoch": 1.86, "learning_rate": 1.907828324847194e-05, "loss": 0.6428, "step": 111500 }, { "epoch": 1.86, "learning_rate": 1.8939620841514415e-05, "loss": 0.6488, "step": 112000 }, { "epoch": 1.87, "learning_rate": 1.880095843455689e-05, "loss": 0.6384, "step": 112500 }, { "epoch": 1.88, "learning_rate": 1.8662296027599367e-05, "loss": 0.6545, "step": 113000 }, { "epoch": 1.89, "learning_rate": 1.852363362064184e-05, "loss": 0.6477, "step": 113500 }, { "epoch": 1.9, "learning_rate": 1.8384971213684317e-05, "loss": 0.6552, "step": 114000 }, { "epoch": 1.91, "learning_rate": 1.8246308806726793e-05, "loss": 0.6497, "step": 114500 }, { "epoch": 1.91, "learning_rate": 1.8107646399769266e-05, "loss": 0.6604, "step": 115000 }, { "epoch": 1.92, "learning_rate": 1.7968983992811743e-05, "loss": 0.6485, "step": 115500 }, { "epoch": 1.93, "learning_rate": 1.783032158585422e-05, "loss": 0.6455, "step": 116000 }, { "epoch": 1.94, "learning_rate": 1.7691659178896692e-05, "loss": 0.6377, "step": 116500 }, { "epoch": 1.95, "learning_rate": 1.755299677193917e-05, "loss": 0.6406, "step": 117000 }, { "epoch": 1.96, "learning_rate": 1.7414334364981642e-05, "loss": 0.6398, "step": 117500 }, { "epoch": 1.96, "learning_rate": 1.7275671958024118e-05, "loss": 0.6443, "step": 118000 }, { "epoch": 1.97, "learning_rate": 1.713700955106659e-05, "loss": 0.6327, "step": 118500 }, { "epoch": 1.98, "learning_rate": 1.6998347144109068e-05, "loss": 0.6404, "step": 119000 }, { "epoch": 1.99, "learning_rate": 1.6859684737151544e-05, "loss": 0.6442, "step": 119500 }, { "epoch": 2.0, "learning_rate": 1.6721022330194017e-05, "loss": 0.6319, "step": 120000 }, { "epoch": 2.01, "learning_rate": 1.6582359923236494e-05, "loss": 0.5924, "step": 120500 }, { "epoch": 2.01, "learning_rate": 1.6443697516278967e-05, "loss": 0.5649, "step": 121000 }, { "epoch": 2.02, "learning_rate": 1.6305035109321443e-05, "loss": 0.5683, "step": 121500 }, { "epoch": 2.03, "learning_rate": 1.616637270236392e-05, "loss": 0.5665, "step": 122000 }, { "epoch": 2.04, "learning_rate": 1.6027710295406392e-05, "loss": 0.5736, "step": 122500 }, { "epoch": 2.05, "learning_rate": 1.588904788844887e-05, "loss": 0.5595, "step": 123000 }, { "epoch": 2.05, "learning_rate": 1.5750385481491342e-05, "loss": 0.5596, "step": 123500 }, { "epoch": 2.06, "learning_rate": 1.561172307453382e-05, "loss": 0.565, "step": 124000 }, { "epoch": 2.07, "learning_rate": 1.547306066757629e-05, "loss": 0.554, "step": 124500 }, { "epoch": 2.08, "learning_rate": 1.5334398260618768e-05, "loss": 0.5606, "step": 125000 }, { "epoch": 2.09, "learning_rate": 1.5195735853661244e-05, "loss": 0.5725, "step": 125500 }, { "epoch": 2.1, "learning_rate": 1.5057073446703717e-05, "loss": 0.5499, "step": 126000 }, { "epoch": 2.1, "learning_rate": 1.4918411039746194e-05, "loss": 0.5634, "step": 126500 }, { "epoch": 2.11, "learning_rate": 1.4779748632788667e-05, "loss": 0.5664, "step": 127000 }, { "epoch": 2.12, "learning_rate": 1.4641086225831143e-05, "loss": 0.5692, "step": 127500 }, { "epoch": 2.13, "learning_rate": 1.4502423818873618e-05, "loss": 0.5705, "step": 128000 }, { "epoch": 2.14, "learning_rate": 1.4363761411916093e-05, "loss": 0.5569, "step": 128500 }, { "epoch": 2.15, "learning_rate": 1.4225099004958569e-05, "loss": 0.5596, "step": 129000 }, { "epoch": 2.15, "learning_rate": 1.4086436598001044e-05, "loss": 0.5492, "step": 129500 }, { "epoch": 2.16, "learning_rate": 1.3947774191043519e-05, "loss": 0.5534, "step": 130000 }, { "epoch": 2.17, "learning_rate": 1.3809111784085993e-05, "loss": 0.5613, "step": 130500 }, { "epoch": 2.18, "learning_rate": 1.367044937712847e-05, "loss": 0.5682, "step": 131000 }, { "epoch": 2.19, "learning_rate": 1.3531786970170943e-05, "loss": 0.5638, "step": 131500 }, { "epoch": 2.2, "learning_rate": 1.339312456321342e-05, "loss": 0.5755, "step": 132000 }, { "epoch": 2.2, "learning_rate": 1.3254462156255896e-05, "loss": 0.5574, "step": 132500 }, { "epoch": 2.21, "learning_rate": 1.3115799749298369e-05, "loss": 0.5691, "step": 133000 }, { "epoch": 2.22, "learning_rate": 1.2977137342340845e-05, "loss": 0.5591, "step": 133500 }, { "epoch": 2.23, "learning_rate": 1.2838474935383318e-05, "loss": 0.5601, "step": 134000 }, { "epoch": 2.24, "learning_rate": 1.2699812528425795e-05, "loss": 0.5558, "step": 134500 }, { "epoch": 2.25, "learning_rate": 1.2561150121468268e-05, "loss": 0.556, "step": 135000 }, { "epoch": 2.25, "learning_rate": 1.2422487714510744e-05, "loss": 0.5487, "step": 135500 }, { "epoch": 2.26, "learning_rate": 1.2283825307553219e-05, "loss": 0.5625, "step": 136000 }, { "epoch": 2.27, "learning_rate": 1.2145162900595694e-05, "loss": 0.5595, "step": 136500 }, { "epoch": 2.28, "learning_rate": 1.2006500493638168e-05, "loss": 0.5566, "step": 137000 }, { "epoch": 2.29, "learning_rate": 1.1867838086680643e-05, "loss": 0.5573, "step": 137500 }, { "epoch": 2.3, "learning_rate": 1.172917567972312e-05, "loss": 0.5679, "step": 138000 }, { "epoch": 2.3, "learning_rate": 1.1590513272765594e-05, "loss": 0.5678, "step": 138500 }, { "epoch": 2.31, "learning_rate": 1.1451850865808069e-05, "loss": 0.5582, "step": 139000 }, { "epoch": 2.32, "learning_rate": 1.1313188458850545e-05, "loss": 0.5558, "step": 139500 }, { "epoch": 2.33, "learning_rate": 1.117452605189302e-05, "loss": 0.5564, "step": 140000 }, { "epoch": 2.34, "learning_rate": 1.1035863644935495e-05, "loss": 0.5551, "step": 140500 }, { "epoch": 2.35, "learning_rate": 1.089720123797797e-05, "loss": 0.5582, "step": 141000 }, { "epoch": 2.35, "learning_rate": 1.0758538831020446e-05, "loss": 0.5558, "step": 141500 }, { "epoch": 2.36, "learning_rate": 1.061987642406292e-05, "loss": 0.5594, "step": 142000 }, { "epoch": 2.37, "learning_rate": 1.0481214017105396e-05, "loss": 0.5576, "step": 142500 }, { "epoch": 2.38, "learning_rate": 1.034255161014787e-05, "loss": 0.5557, "step": 143000 }, { "epoch": 2.39, "learning_rate": 1.0203889203190345e-05, "loss": 0.5567, "step": 143500 }, { "epoch": 2.4, "learning_rate": 1.006522679623282e-05, "loss": 0.5592, "step": 144000 }, { "epoch": 2.4, "learning_rate": 9.926564389275294e-06, "loss": 0.5555, "step": 144500 }, { "epoch": 2.41, "learning_rate": 9.787901982317771e-06, "loss": 0.5528, "step": 145000 }, { "epoch": 2.42, "learning_rate": 9.649239575360246e-06, "loss": 0.5496, "step": 145500 }, { "epoch": 2.43, "learning_rate": 9.51057716840272e-06, "loss": 0.5588, "step": 146000 }, { "epoch": 2.44, "learning_rate": 9.371914761445195e-06, "loss": 0.5559, "step": 146500 }, { "epoch": 2.45, "learning_rate": 9.23325235448767e-06, "loss": 0.5604, "step": 147000 }, { "epoch": 2.45, "learning_rate": 9.094589947530145e-06, "loss": 0.5542, "step": 147500 }, { "epoch": 2.46, "learning_rate": 8.95592754057262e-06, "loss": 0.5512, "step": 148000 }, { "epoch": 2.47, "learning_rate": 8.817265133615096e-06, "loss": 0.5521, "step": 148500 }, { "epoch": 2.48, "learning_rate": 8.67860272665757e-06, "loss": 0.5523, "step": 149000 }, { "epoch": 2.49, "learning_rate": 8.539940319700045e-06, "loss": 0.5443, "step": 149500 }, { "epoch": 2.5, "learning_rate": 8.401277912742522e-06, "loss": 0.5555, "step": 150000 }, { "epoch": 2.5, "learning_rate": 8.262615505784996e-06, "loss": 0.5434, "step": 150500 }, { "epoch": 2.51, "learning_rate": 8.123953098827471e-06, "loss": 0.5527, "step": 151000 }, { "epoch": 2.52, "learning_rate": 7.985290691869946e-06, "loss": 0.5503, "step": 151500 }, { "epoch": 2.53, "learning_rate": 7.846628284912422e-06, "loss": 0.5496, "step": 152000 }, { "epoch": 2.54, "learning_rate": 7.707965877954897e-06, "loss": 0.5401, "step": 152500 }, { "epoch": 2.55, "learning_rate": 7.569303470997372e-06, "loss": 0.5506, "step": 153000 }, { "epoch": 2.55, "learning_rate": 7.4306410640398465e-06, "loss": 0.557, "step": 153500 }, { "epoch": 2.56, "learning_rate": 7.291978657082321e-06, "loss": 0.5462, "step": 154000 }, { "epoch": 2.57, "learning_rate": 7.153316250124796e-06, "loss": 0.5561, "step": 154500 }, { "epoch": 2.58, "learning_rate": 7.014653843167271e-06, "loss": 0.5479, "step": 155000 }, { "epoch": 2.59, "learning_rate": 6.875991436209747e-06, "loss": 0.551, "step": 155500 }, { "epoch": 2.6, "learning_rate": 6.737329029252222e-06, "loss": 0.5445, "step": 156000 }, { "epoch": 2.6, "learning_rate": 6.598666622294697e-06, "loss": 0.5399, "step": 156500 }, { "epoch": 2.61, "learning_rate": 6.460004215337171e-06, "loss": 0.5511, "step": 157000 }, { "epoch": 2.62, "learning_rate": 6.321341808379647e-06, "loss": 0.5509, "step": 157500 }, { "epoch": 2.63, "learning_rate": 6.1826794014221225e-06, "loss": 0.5478, "step": 158000 }, { "epoch": 2.64, "learning_rate": 6.044016994464597e-06, "loss": 0.5434, "step": 158500 }, { "epoch": 2.65, "learning_rate": 5.905354587507072e-06, "loss": 0.548, "step": 159000 }, { "epoch": 2.65, "learning_rate": 5.7666921805495476e-06, "loss": 0.5613, "step": 159500 }, { "epoch": 2.66, "learning_rate": 5.628029773592022e-06, "loss": 0.5487, "step": 160000 }, { "epoch": 2.67, "learning_rate": 5.489367366634497e-06, "loss": 0.551, "step": 160500 }, { "epoch": 2.68, "learning_rate": 5.350704959676972e-06, "loss": 0.5485, "step": 161000 }, { "epoch": 2.69, "learning_rate": 5.212042552719447e-06, "loss": 0.5433, "step": 161500 }, { "epoch": 2.7, "learning_rate": 5.073380145761922e-06, "loss": 0.5423, "step": 162000 }, { "epoch": 2.7, "learning_rate": 4.934717738804398e-06, "loss": 0.5475, "step": 162500 }, { "epoch": 2.71, "learning_rate": 4.796055331846873e-06, "loss": 0.5441, "step": 163000 }, { "epoch": 2.72, "learning_rate": 4.657392924889348e-06, "loss": 0.5592, "step": 163500 }, { "epoch": 2.73, "learning_rate": 4.518730517931823e-06, "loss": 0.5347, "step": 164000 }, { "epoch": 2.74, "learning_rate": 4.3800681109742975e-06, "loss": 0.523, "step": 164500 }, { "epoch": 2.75, "learning_rate": 4.241405704016773e-06, "loss": 0.5407, "step": 165000 }, { "epoch": 2.75, "learning_rate": 4.102743297059248e-06, "loss": 0.5391, "step": 165500 }, { "epoch": 2.76, "learning_rate": 3.9640808901017225e-06, "loss": 0.5403, "step": 166000 }, { "epoch": 2.77, "learning_rate": 3.825418483144198e-06, "loss": 0.5397, "step": 166500 }, { "epoch": 2.78, "learning_rate": 3.6867560761866733e-06, "loss": 0.5446, "step": 167000 }, { "epoch": 2.79, "learning_rate": 3.548093669229148e-06, "loss": 0.5422, "step": 167500 }, { "epoch": 2.8, "learning_rate": 3.4094312622716227e-06, "loss": 0.5428, "step": 168000 }, { "epoch": 2.8, "learning_rate": 3.2707688553140983e-06, "loss": 0.5433, "step": 168500 }, { "epoch": 2.81, "learning_rate": 3.1321064483565735e-06, "loss": 0.5235, "step": 169000 }, { "epoch": 2.82, "learning_rate": 2.9934440413990486e-06, "loss": 0.5394, "step": 169500 }, { "epoch": 2.83, "learning_rate": 2.8547816344415234e-06, "loss": 0.5416, "step": 170000 }, { "epoch": 2.84, "learning_rate": 2.7161192274839985e-06, "loss": 0.5407, "step": 170500 }, { "epoch": 2.85, "learning_rate": 2.5774568205264733e-06, "loss": 0.5323, "step": 171000 }, { "epoch": 2.85, "learning_rate": 2.438794413568949e-06, "loss": 0.5463, "step": 171500 }, { "epoch": 2.86, "learning_rate": 2.3001320066114236e-06, "loss": 0.5348, "step": 172000 }, { "epoch": 2.87, "learning_rate": 2.1614695996538988e-06, "loss": 0.5341, "step": 172500 }, { "epoch": 2.88, "learning_rate": 2.022807192696374e-06, "loss": 0.54, "step": 173000 }, { "epoch": 2.89, "learning_rate": 1.8841447857388489e-06, "loss": 0.5435, "step": 173500 }, { "epoch": 2.9, "learning_rate": 1.745482378781324e-06, "loss": 0.5384, "step": 174000 }, { "epoch": 2.9, "learning_rate": 1.606819971823799e-06, "loss": 0.5398, "step": 174500 }, { "epoch": 2.91, "learning_rate": 1.4681575648662741e-06, "loss": 0.539, "step": 175000 }, { "epoch": 2.92, "learning_rate": 1.329495157908749e-06, "loss": 0.5348, "step": 175500 }, { "epoch": 2.93, "learning_rate": 1.1908327509512242e-06, "loss": 0.5231, "step": 176000 }, { "epoch": 2.94, "learning_rate": 1.0521703439936992e-06, "loss": 0.5353, "step": 176500 }, { "epoch": 2.95, "learning_rate": 9.135079370361742e-07, "loss": 0.5421, "step": 177000 }, { "epoch": 2.95, "learning_rate": 7.748455300786494e-07, "loss": 0.5531, "step": 177500 }, { "epoch": 2.96, "learning_rate": 6.361831231211245e-07, "loss": 0.5364, "step": 178000 }, { "epoch": 2.97, "learning_rate": 4.975207161635994e-07, "loss": 0.5406, "step": 178500 }, { "epoch": 2.98, "learning_rate": 3.588583092060745e-07, "loss": 0.5395, "step": 179000 }, { "epoch": 2.99, "learning_rate": 2.201959022485496e-07, "loss": 0.5286, "step": 179500 }, { "epoch": 3.0, "learning_rate": 8.153349529102467e-08, "loss": 0.5469, "step": 180000 }, { "epoch": 3.0, "step": 180294, "total_flos": 2.0490419005872538e+17, "train_loss": 0.45402886773448525, "train_runtime": 39387.4889, "train_samples_per_second": 45.774, "train_steps_per_second": 4.577 } ], "max_steps": 180294, "num_train_epochs": 3, "total_flos": 2.0490419005872538e+17, "trial_name": null, "trial_params": null }