End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8774193548387097,
-    "eval_loss": 0.44974660873413086,
-    "eval_runtime": 14.3146,
-    "eval_samples_per_second": 10.828,
-    "eval_steps_per_second": 1.397
+    "eval_accuracy": 0.8516129032258064,
+    "eval_loss": 0.40030112862586975,
+    "eval_runtime": 12.2214,
+    "eval_samples_per_second": 12.683,
+    "eval_steps_per_second": 1.636
 }

runs/Oct14_14-25-55_71fb6b0700bc/events.out.tfevents.1728916191.71fb6b0700bc.3447.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e4d376c0d63ff416733d2b02ead7afe40ca7595daa9f82900f83b8ef66b0a1eb
-size 411
+oid sha256:7ed45c7bb655f1b9eccd5de2dbfeedfd8c8edfe3b841250e18ca648f19b3fdec
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8774193548387097,
-    "eval_loss": 0.44974660873413086,
-    "eval_runtime": 14.3146,
-    "eval_samples_per_second": 10.828,
-    "eval_steps_per_second": 1.397
+    "eval_accuracy": 0.8516129032258064,
+    "eval_loss": 0.40030112862586975,
+    "eval_runtime": 12.2214,
+    "eval_samples_per_second": 12.683,
+    "eval_steps_per_second": 1.636
 }

trainer_state.json

@@ -10,163 +10,163 @@
   "log_history": [
     {
       "epoch": 0.06756756756756757,
-      "grad_norm": 6.423978805541992,
+      "grad_norm": 6.236786842346191,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.3465,
+      "loss": 2.3922,
       "step": 10
     },
     {
       "epoch": 0.13513513513513514,
-      "grad_norm": 7.651690483093262,
+      "grad_norm": 6.904232978820801,
       "learning_rate": 4.81203007518797e-05,
-      "loss": 2.179,
+      "loss": 2.2012,
       "step": 20
     },
     {
       "epoch": 0.20270270270270271,
-      "grad_norm": 7.18040657043457,
+      "grad_norm": 7.755519390106201,
       "learning_rate": 4.43609022556391e-05,
-      "loss": 2.1243,
+      "loss": 2.1814,
       "step": 30
     },
     {
       "epoch": 0.25675675675675674,
-      "eval_accuracy": 0.38571428571428573,
-      "eval_loss": 1.8105360269546509,
-      "eval_runtime": 6.708,
-      "eval_samples_per_second": 10.435,
-      "eval_steps_per_second": 1.342,
+      "eval_accuracy": 0.4714285714285714,
+      "eval_loss": 1.857365369796753,
+      "eval_runtime": 6.4158,
+      "eval_samples_per_second": 10.911,
+      "eval_steps_per_second": 1.403,
       "step": 38
     },
     {
       "epoch": 1.0135135135135136,
-      "grad_norm": 9.158327102661133,
+      "grad_norm": 8.18625259399414,
       "learning_rate": 4.0601503759398494e-05,
-      "loss": 1.9503,
+      "loss": 2.0015,
       "step": 40
     },
     {
       "epoch": 1.0810810810810811,
-      "grad_norm": 10.240347862243652,
+      "grad_norm": 11.397420883178711,
       "learning_rate": 3.6842105263157895e-05,
-      "loss": 1.5053,
+      "loss": 1.6643,
       "step": 50
     },
     {
       "epoch": 1.1486486486486487,
-      "grad_norm": 9.871286392211914,
+      "grad_norm": 11.84655475616455,
       "learning_rate": 3.3082706766917295e-05,
-      "loss": 1.2446,
+      "loss": 1.2146,
       "step": 60
     },
     {
       "epoch": 1.2162162162162162,
-      "grad_norm": 7.637844085693359,
+      "grad_norm": 6.57381534576416,
       "learning_rate": 2.9323308270676693e-05,
-      "loss": 0.9539,
+      "loss": 0.8234,
       "step": 70
     },
     {
       "epoch": 1.2567567567567568,
-      "eval_accuracy": 0.5857142857142857,
-      "eval_loss": 0.9814803004264832,
-      "eval_runtime": 6.536,
-      "eval_samples_per_second": 10.71,
-      "eval_steps_per_second": 1.377,
+      "eval_accuracy": 0.7,
+      "eval_loss": 0.9412073493003845,
+      "eval_runtime": 6.491,
+      "eval_samples_per_second": 10.784,
+      "eval_steps_per_second": 1.387,
       "step": 76
     },
     {
       "epoch": 2.027027027027027,
-      "grad_norm": 5.889892101287842,
+      "grad_norm": 7.174142837524414,
       "learning_rate": 2.556390977443609e-05,
-      "loss": 0.8534,
+      "loss": 0.8626,
       "step": 80
     },
     {
       "epoch": 2.0945945945945947,
-      "grad_norm": 14.065513610839844,
+      "grad_norm": 9.910486221313477,
       "learning_rate": 2.1804511278195487e-05,
-      "loss": 0.6305,
+      "loss": 0.6279,
       "step": 90
     },
     {
       "epoch": 2.1621621621621623,
-      "grad_norm": 8.854910850524902,
+      "grad_norm": 18.203344345092773,
       "learning_rate": 1.8045112781954888e-05,
-      "loss": 0.5601,
+      "loss": 0.5971,
       "step": 100
     },
     {
       "epoch": 2.22972972972973,
-      "grad_norm": 7.27897834777832,
+      "grad_norm": 5.582463264465332,
       "learning_rate": 1.4285714285714285e-05,
-      "loss": 0.4701,
+      "loss": 0.467,
       "step": 110
     },
     {
       "epoch": 2.2567567567567566,
-      "eval_accuracy": 0.8571428571428571,
-      "eval_loss": 0.48053404688835144,
-      "eval_runtime": 6.7679,
-      "eval_samples_per_second": 10.343,
-      "eval_steps_per_second": 1.33,
+      "eval_accuracy": 0.7571428571428571,
+      "eval_loss": 0.5703738331794739,
+      "eval_runtime": 6.3262,
+      "eval_samples_per_second": 11.065,
+      "eval_steps_per_second": 1.423,
       "step": 114
     },
     {
       "epoch": 3.0405405405405403,
-      "grad_norm": 3.358264207839966,
+      "grad_norm": 3.68727707862854,
       "learning_rate": 1.0526315789473684e-05,
-      "loss": 0.3716,
+      "loss": 0.4091,
       "step": 120
     },
     {
       "epoch": 3.108108108108108,
-      "grad_norm": 3.2281341552734375,
+      "grad_norm": 3.5390589237213135,
       "learning_rate": 6.766917293233083e-06,
-      "loss": 0.3068,
+      "loss": 0.2842,
       "step": 130
     },
     {
       "epoch": 3.175675675675676,
-      "grad_norm": 4.061362266540527,
+      "grad_norm": 8.368684768676758,
       "learning_rate": 3.007518796992481e-06,
-      "loss": 0.2869,
+      "loss": 0.3374,
       "step": 140
     },
     {
       "epoch": 3.22972972972973,
       "eval_accuracy": 0.8714285714285714,
-      "eval_loss": 0.39165422320365906,
-      "eval_runtime": 6.849,
-      "eval_samples_per_second": 10.22,
-      "eval_steps_per_second": 1.314,
+      "eval_loss": 0.42163416743278503,
+      "eval_runtime": 6.6066,
+      "eval_samples_per_second": 10.595,
+      "eval_steps_per_second": 1.362,
       "step": 148
     },
     {
       "epoch": 3.22972972972973,
       "step": 148,
       "total_flos": 1.460491890402263e+18,
-      "train_loss": 1.084407622749741,
-      "train_runtime": 223.0988,
-      "train_samples_per_second": 5.307,
-      "train_steps_per_second": 0.663
+      "train_loss": 1.0991543885823842,
+      "train_runtime": 217.7469,
+      "train_samples_per_second": 5.438,
+      "train_steps_per_second": 0.68
     },
     {
       "epoch": 3.22972972972973,
-      "eval_accuracy": 0.8774193548387097,
-      "eval_loss": 0.44974666833877563,
-      "eval_runtime": 14.875,
-      "eval_samples_per_second": 10.42,
-      "eval_steps_per_second": 1.345,
+      "eval_accuracy": 0.8516129032258064,
+      "eval_loss": 0.40030112862586975,
+      "eval_runtime": 13.2529,
+      "eval_samples_per_second": 11.696,
+      "eval_steps_per_second": 1.509,
       "step": 148
     },
     {
       "epoch": 3.22972972972973,
-      "eval_accuracy": 0.8774193548387097,
-      "eval_loss": 0.44974660873413086,
-      "eval_runtime": 14.3146,
-      "eval_samples_per_second": 10.828,
-      "eval_steps_per_second": 1.397,
+      "eval_accuracy": 0.8516129032258064,
+      "eval_loss": 0.40030112862586975,
+      "eval_runtime": 12.2214,
+      "eval_samples_per_second": 12.683,
+      "eval_steps_per_second": 1.636,
       "step": 148
     }
   ],