End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8129032258064516,
-    "eval_loss": 0.5470468997955322,
-    "eval_runtime": 14.4978,
-    "eval_samples_per_second": 10.691,
-    "eval_steps_per_second": 1.38
+    "eval_accuracy": 0.8774193548387097,
+    "eval_loss": 0.44974660873413086,
+    "eval_runtime": 14.3146,
+    "eval_samples_per_second": 10.828,
+    "eval_steps_per_second": 1.397
 }

runs/Oct14_13-48-17_dc6088e9f91b/events.out.tfevents.1728913945.dc6088e9f91b.3595.3

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:29740312ac3114f4330945689b71d7f47ca17aa25fe564f86bcb35fcfb0571ff
-size 411
+oid sha256:6307ae85a02605a5fc1ddd68f49983f36817f22948a7500f1096bc73bbd8cf27
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8129032258064516,
-    "eval_loss": 0.5470468997955322,
-    "eval_runtime": 14.4978,
-    "eval_samples_per_second": 10.691,
-    "eval_steps_per_second": 1.38
+    "eval_accuracy": 0.8774193548387097,
+    "eval_loss": 0.44974660873413086,
+    "eval_runtime": 14.3146,
+    "eval_samples_per_second": 10.828,
+    "eval_steps_per_second": 1.397
 }

trainer_state.json

@@ -10,163 +10,163 @@
   "log_history": [
     {
       "epoch": 0.06756756756756757,
-      "grad_norm": 6.49423885345459,
+      "grad_norm": 6.423978805541992,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.3757,
+      "loss": 2.3465,
       "step": 10
     },
     {
       "epoch": 0.13513513513513514,
-      "grad_norm": 6.615733623504639,
+      "grad_norm": 7.651690483093262,
       "learning_rate": 4.81203007518797e-05,
-      "loss": 2.2277,
+      "loss": 2.179,
       "step": 20
     },
     {
       "epoch": 0.20270270270270271,
-      "grad_norm": 8.014123916625977,
+      "grad_norm": 7.18040657043457,
       "learning_rate": 4.43609022556391e-05,
-      "loss": 2.109,
+      "loss": 2.1243,
       "step": 30
     },
     {
       "epoch": 0.25675675675675674,
-      "eval_accuracy": 0.6,
-      "eval_loss": 1.8528673648834229,
-      "eval_runtime": 6.7471,
-      "eval_samples_per_second": 10.375,
-      "eval_steps_per_second": 1.334,
+      "eval_accuracy": 0.38571428571428573,
+      "eval_loss": 1.8105360269546509,
+      "eval_runtime": 6.708,
+      "eval_samples_per_second": 10.435,
+      "eval_steps_per_second": 1.342,
       "step": 38
     },
     {
       "epoch": 1.0135135135135136,
-      "grad_norm": 14.876914978027344,
+      "grad_norm": 9.158327102661133,
       "learning_rate": 4.0601503759398494e-05,
-      "loss": 1.9624,
+      "loss": 1.9503,
       "step": 40
     },
     {
       "epoch": 1.0810810810810811,
-      "grad_norm": 11.556279182434082,
+      "grad_norm": 10.240347862243652,
       "learning_rate": 3.6842105263157895e-05,
-      "loss": 1.5381,
+      "loss": 1.5053,
       "step": 50
     },
     {
       "epoch": 1.1486486486486487,
-      "grad_norm": 12.52138900756836,
+      "grad_norm": 9.871286392211914,
       "learning_rate": 3.3082706766917295e-05,
-      "loss": 1.2444,
+      "loss": 1.2446,
       "step": 60
     },
     {
       "epoch": 1.2162162162162162,
-      "grad_norm": 5.785777568817139,
+      "grad_norm": 7.637844085693359,
       "learning_rate": 2.9323308270676693e-05,
-      "loss": 0.9259,
+      "loss": 0.9539,
       "step": 70
     },
     {
       "epoch": 1.2567567567567568,
-      "eval_accuracy": 0.5,
-      "eval_loss": 1.2051961421966553,
-      "eval_runtime": 6.2919,
-      "eval_samples_per_second": 11.125,
-      "eval_steps_per_second": 1.43,
+      "eval_accuracy": 0.5857142857142857,
+      "eval_loss": 0.9814803004264832,
+      "eval_runtime": 6.536,
+      "eval_samples_per_second": 10.71,
+      "eval_steps_per_second": 1.377,
       "step": 76
     },
     {
       "epoch": 2.027027027027027,
-      "grad_norm": 6.8339762687683105,
+      "grad_norm": 5.889892101287842,
       "learning_rate": 2.556390977443609e-05,
-      "loss": 0.8577,
+      "loss": 0.8534,
       "step": 80
     },
     {
       "epoch": 2.0945945945945947,
-      "grad_norm": 11.754122734069824,
+      "grad_norm": 14.065513610839844,
       "learning_rate": 2.1804511278195487e-05,
-      "loss": 0.6724,
+      "loss": 0.6305,
       "step": 90
     },
     {
       "epoch": 2.1621621621621623,
-      "grad_norm": 9.548654556274414,
+      "grad_norm": 8.854910850524902,
       "learning_rate": 1.8045112781954888e-05,
-      "loss": 0.6248,
+      "loss": 0.5601,
       "step": 100
     },
     {
       "epoch": 2.22972972972973,
-      "grad_norm": 7.923789024353027,
+      "grad_norm": 7.27897834777832,
       "learning_rate": 1.4285714285714285e-05,
-      "loss": 0.4802,
+      "loss": 0.4701,
       "step": 110
     },
     {
       "epoch": 2.2567567567567566,
       "eval_accuracy": 0.8571428571428571,
-      "eval_loss": 0.6029335856437683,
-      "eval_runtime": 6.1746,
-      "eval_samples_per_second": 11.337,
-      "eval_steps_per_second": 1.458,
+      "eval_loss": 0.48053404688835144,
+      "eval_runtime": 6.7679,
+      "eval_samples_per_second": 10.343,
+      "eval_steps_per_second": 1.33,
       "step": 114
     },
     {
       "epoch": 3.0405405405405403,
-      "grad_norm": 4.298054218292236,
+      "grad_norm": 3.358264207839966,
       "learning_rate": 1.0526315789473684e-05,
-      "loss": 0.3842,
+      "loss": 0.3716,
       "step": 120
     },
     {
       "epoch": 3.108108108108108,
-      "grad_norm": 3.788994312286377,
+      "grad_norm": 3.2281341552734375,
       "learning_rate": 6.766917293233083e-06,
-      "loss": 0.2912,
+      "loss": 0.3068,
       "step": 130
     },
     {
       "epoch": 3.175675675675676,
-      "grad_norm": 12.958038330078125,
+      "grad_norm": 4.061362266540527,
       "learning_rate": 3.007518796992481e-06,
-      "loss": 0.3519,
+      "loss": 0.2869,
       "step": 140
     },
     {
       "epoch": 3.22972972972973,
       "eval_accuracy": 0.8714285714285714,
-      "eval_loss": 0.4826556146144867,
-      "eval_runtime": 6.6956,
-      "eval_samples_per_second": 10.455,
-      "eval_steps_per_second": 1.344,
+      "eval_loss": 0.39165422320365906,
+      "eval_runtime": 6.849,
+      "eval_samples_per_second": 10.22,
+      "eval_steps_per_second": 1.314,
       "step": 148
     },
     {
       "epoch": 3.22972972972973,
       "step": 148,
       "total_flos": 1.460491890402263e+18,
-      "train_loss": 1.1000467377739984,
-      "train_runtime": 221.0945,
-      "train_samples_per_second": 5.355,
-      "train_steps_per_second": 0.669
+      "train_loss": 1.084407622749741,
+      "train_runtime": 223.0988,
+      "train_samples_per_second": 5.307,
+      "train_steps_per_second": 0.663
     },
     {
       "epoch": 3.22972972972973,
-      "eval_accuracy": 0.8129032258064516,
-      "eval_loss": 0.5470468401908875,
-      "eval_runtime": 14.5546,
-      "eval_samples_per_second": 10.65,
-      "eval_steps_per_second": 1.374,
+      "eval_accuracy": 0.8774193548387097,
+      "eval_loss": 0.44974666833877563,
+      "eval_runtime": 14.875,
+      "eval_samples_per_second": 10.42,
+      "eval_steps_per_second": 1.345,
       "step": 148
     },
     {
       "epoch": 3.22972972972973,
-      "eval_accuracy": 0.8129032258064516,
-      "eval_loss": 0.5470468997955322,
-      "eval_runtime": 14.4978,
-      "eval_samples_per_second": 10.691,
-      "eval_steps_per_second": 1.38,
+      "eval_accuracy": 0.8774193548387097,
+      "eval_loss": 0.44974660873413086,
+      "eval_runtime": 14.3146,
+      "eval_samples_per_second": 10.828,
+      "eval_steps_per_second": 1.397,
       "step": 148
     }
   ],