Training in progress, step 400, checkpoint

checkpoint-400/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e163f1e34bea240e47931e35bddb3af98e1a17942fed17912121d0e0c8cbc6cb
+oid sha256:9a08500d0882f7ef489b6fc38b7ca922af152fede8ad1be8a687897a93c68fca
 size 9873829

checkpoint-400/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e50caa3eb6cc267f4f35facb542e79137e73880bdf78c5bd9855d0f11fbeb62e
+oid sha256:2765f7ecc3c9853b8cd6f6b5619b326032c6eb89d56c5aca981e370accd23d1b
 size 42788

checkpoint-400/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ce8f53cf3661e7021586919d9382d8320c5ef41aa71edeb37eba6cdf90db5eb5
+oid sha256:94cf04def08d1529bbdde5d3ecd3383aaa07e7e13d40d9bff68b23a599407791
 size 14244

checkpoint-400/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0af3670fec8fd4d7affcd6c02fd14d5dd405e87af8e413b3f19b6f442abe5b5e
+oid sha256:1b12f69cba539ef7162900377a9e7c168038d643a5ca8ec3f033fd12ca09a416
 size 1064

checkpoint-400/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.665068507194519,
-  "best_model_checkpoint": "output/checkpoint-50",
-  "epoch": 1.6368286445012787,
+  "best_metric": 0.5725014805793762,
+  "best_model_checkpoint": "output/checkpoint-100",
+  "epoch": 7.547169811320755,
   "eval_steps": 50,
   "global_step": 400,
   "is_hyper_param_search": false,
@@ -9,137 +9,137 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0,
-      "learning_rate": 7.575757575757576e-07,
-      "loss": 0.4968,
+      "epoch": 0.02,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.4437,
       "step": 1
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 3.484848484848485e-05,
-      "loss": 0.5254,
+      "epoch": 0.94,
+      "learning_rate": 5.679012345679012e-05,
+      "loss": 0.455,
       "step": 50
     },
     {
-      "epoch": 0.2,
-      "eval_accuracy": 0.6507462686567164,
-      "eval_loss": 0.665068507194519,
-      "eval_runtime": 61.5592,
-      "eval_samples_per_second": 5.442,
-      "eval_steps_per_second": 1.365,
+      "epoch": 0.94,
+      "eval_accuracy": 0.68,
+      "eval_loss": 0.6171127557754517,
+      "eval_runtime": 29.0494,
+      "eval_samples_per_second": 6.024,
+      "eval_steps_per_second": 1.515,
       "step": 50
     },
     {
-      "epoch": 0.41,
-      "learning_rate": 7.272727272727273e-05,
-      "loss": 0.4971,
+      "epoch": 1.89,
+      "learning_rate": 9.59349593495935e-05,
+      "loss": 0.4555,
       "step": 100
     },
     {
-      "epoch": 0.41,
-      "eval_accuracy": 0.6656716417910448,
-      "eval_loss": 0.8002150058746338,
-      "eval_runtime": 61.211,
-      "eval_samples_per_second": 5.473,
-      "eval_steps_per_second": 1.372,
+      "epoch": 1.89,
+      "eval_accuracy": 0.7257142857142858,
+      "eval_loss": 0.5725014805793762,
+      "eval_runtime": 28.7986,
+      "eval_samples_per_second": 6.077,
+      "eval_steps_per_second": 1.528,
       "step": 100
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 9.783333333333334e-05,
-      "loss": 0.5039,
+      "epoch": 2.83,
+      "learning_rate": 8.265582655826559e-05,
+      "loss": 0.455,
       "step": 150
     },
     {
-      "epoch": 0.61,
-      "eval_accuracy": 0.6865671641791045,
-      "eval_loss": 0.7404947280883789,
-      "eval_runtime": 61.4487,
-      "eval_samples_per_second": 5.452,
-      "eval_steps_per_second": 1.367,
+      "epoch": 2.83,
+      "eval_accuracy": 0.7371428571428571,
+      "eval_loss": 0.5729417204856873,
+      "eval_runtime": 28.9054,
+      "eval_samples_per_second": 6.054,
+      "eval_steps_per_second": 1.522,
       "step": 150
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 8.950000000000001e-05,
-      "loss": 0.4944,
+      "epoch": 3.77,
+      "learning_rate": 6.910569105691057e-05,
+      "loss": 0.4718,
       "step": 200
     },
     {
-      "epoch": 0.82,
-      "eval_accuracy": 0.6238805970149254,
-      "eval_loss": 0.9986834526062012,
-      "eval_runtime": 60.5135,
-      "eval_samples_per_second": 5.536,
-      "eval_steps_per_second": 1.388,
+      "epoch": 3.77,
+      "eval_accuracy": 0.6685714285714286,
+      "eval_loss": 0.6322054862976074,
+      "eval_runtime": 28.8878,
+      "eval_samples_per_second": 6.058,
+      "eval_steps_per_second": 1.523,
       "step": 200
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 8.116666666666667e-05,
-      "loss": 0.508,
+      "epoch": 4.72,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.4348,
       "step": 250
     },
     {
-      "epoch": 1.02,
-      "eval_accuracy": 0.6268656716417911,
-      "eval_loss": 0.9990125894546509,
-      "eval_runtime": 56.4381,
-      "eval_samples_per_second": 5.936,
-      "eval_steps_per_second": 1.488,
+      "epoch": 4.72,
+      "eval_accuracy": 0.7257142857142858,
+      "eval_loss": 0.5733612179756165,
+      "eval_runtime": 28.5713,
+      "eval_samples_per_second": 6.125,
+      "eval_steps_per_second": 1.54,
       "step": 250
     },
     {
-      "epoch": 1.23,
-      "learning_rate": 7.283333333333335e-05,
-      "loss": 0.4904,
+      "epoch": 5.66,
+      "learning_rate": 4.2005420054200545e-05,
+      "loss": 0.4353,
       "step": 300
     },
     {
-      "epoch": 1.23,
-      "eval_accuracy": 0.6119402985074627,
-      "eval_loss": 1.0483125448226929,
-      "eval_runtime": 56.4117,
-      "eval_samples_per_second": 5.938,
-      "eval_steps_per_second": 1.489,
+      "epoch": 5.66,
+      "eval_accuracy": 0.6628571428571428,
+      "eval_loss": 0.6271094679832458,
+      "eval_runtime": 26.5252,
+      "eval_samples_per_second": 6.597,
+      "eval_steps_per_second": 1.659,
       "step": 300
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 6.450000000000001e-05,
-      "loss": 0.4694,
+      "epoch": 6.6,
+      "learning_rate": 2.8455284552845528e-05,
+      "loss": 0.4261,
       "step": 350
     },
     {
-      "epoch": 1.43,
-      "eval_accuracy": 0.6417910447761194,
-      "eval_loss": 0.9735069274902344,
-      "eval_runtime": 56.422,
-      "eval_samples_per_second": 5.937,
-      "eval_steps_per_second": 1.489,
+      "epoch": 6.6,
+      "eval_accuracy": 0.6685714285714286,
+      "eval_loss": 0.6181718111038208,
+      "eval_runtime": 26.5098,
+      "eval_samples_per_second": 6.601,
+      "eval_steps_per_second": 1.66,
       "step": 350
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 5.6166666666666665e-05,
-      "loss": 0.4937,
+      "epoch": 7.55,
+      "learning_rate": 1.4905149051490516e-05,
+      "loss": 0.4213,
       "step": 400
     },
     {
-      "epoch": 1.64,
-      "eval_accuracy": 0.5880597014925373,
-      "eval_loss": 1.1042370796203613,
-      "eval_runtime": 56.4248,
-      "eval_samples_per_second": 5.937,
-      "eval_steps_per_second": 1.489,
+      "epoch": 7.55,
+      "eval_accuracy": 0.6628571428571428,
+      "eval_loss": 0.6409146785736084,
+      "eval_runtime": 26.5165,
+      "eval_samples_per_second": 6.6,
+      "eval_steps_per_second": 1.659,
       "step": 400
     }
   ],
   "logging_steps": 50,
-  "max_steps": 732,
-  "num_train_epochs": 3,
+  "max_steps": 450,
+  "num_train_epochs": 9,
   "save_steps": 50,
-  "total_flos": 1.7066190319742976e+17,
+  "total_flos": 1.5313215392907264e+17,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-400/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9feea516a8a208c393c324d84d047c0ab5036bbdf3fb1f3eea6a2563e9c79a6b
+oid sha256:d91c1e3ec7fc28d5ed2a9d783b7ded0a7affbbd029319e299d9244ac0f3fbc29
 size 4472