Training in progress, step 500, checkpoint

checkpoint-500/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ddbec630d8d4e2ba6cfe8de48d4c8c1f5f2dcf6f31eb064b794a1f3a7a1ec62a
+oid sha256:2b2842243bc76857104a9e5b962291ed2dd00fd60b178b57f459df32144fee91
 size 9873829

checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:32ef827479040b3af3cc83b91a92d9cbaae2a335d6970a9615f303092f3f68ad
+oid sha256:9792e339d4f329dc29a8b8685ef8f386e609ddbc3f2f21c7557f124bca412af8
 size 42788

checkpoint-500/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d68fe56cbf15026758f37c8a680b1535157bd18583612558f095fc58469cc5c
+oid sha256:0d04af85a3458340d9d25c49454f2697c9d59a5e63020c12b442bda805340d24
 size 14244

checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cea8b9d0f050de9b969bba9b4cb43afde7581d6291473c7c1b729a06cb63c7a2
+oid sha256:ce30089cd6890b74bbc903cba8a59f76da93d625d6d5e187c087192d2e2fb830
 size 1064

checkpoint-500/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.6430063247680664,
+  "best_metric": 0.665068507194519,
   "best_model_checkpoint": "output/checkpoint-50",
-  "epoch": 0.4008417677121956,
+  "epoch": 2.0460358056265986,
   "eval_steps": 50,
   "global_step": 500,
   "is_hyper_param_search": false,
@@ -10,166 +10,166 @@
   "log_history": [
     {
       "epoch": 0.0,
-      "learning_rate": 4.444444444444445e-07,
-      "loss": 0.5542,
+      "learning_rate": 7.575757575757576e-07,
+      "loss": 0.4968,
       "step": 1
     },
     {
-      "epoch": 0.04,
-      "learning_rate": 2e-05,
-      "loss": 0.554,
+      "epoch": 0.2,
+      "learning_rate": 3.484848484848485e-05,
+      "loss": 0.5254,
       "step": 50
     },
     {
-      "epoch": 0.04,
-      "eval_accuracy": 0.6626865671641791,
-      "eval_loss": 0.6430063247680664,
-      "eval_runtime": 61.7692,
-      "eval_samples_per_second": 5.423,
-      "eval_steps_per_second": 1.36,
+      "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,
       "step": 50
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 4.222222222222222e-05,
-      "loss": 0.5957,
+      "epoch": 0.41,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 0.4971,
       "step": 100
     },
     {
-      "epoch": 0.08,
-      "eval_accuracy": 0.6686567164179105,
-      "eval_loss": 0.6705919504165649,
-      "eval_runtime": 60.4388,
-      "eval_samples_per_second": 5.543,
-      "eval_steps_per_second": 1.39,
+      "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,
       "step": 100
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 6.444444444444446e-05,
-      "loss": 0.5668,
+      "epoch": 0.61,
+      "learning_rate": 9.783333333333334e-05,
+      "loss": 0.5039,
       "step": 150
     },
     {
-      "epoch": 0.12,
-      "eval_accuracy": 0.6805970149253732,
-      "eval_loss": 0.7224913835525513,
-      "eval_runtime": 56.4882,
-      "eval_samples_per_second": 5.93,
-      "eval_steps_per_second": 1.487,
+      "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,
       "step": 150
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 8.666666666666667e-05,
-      "loss": 0.5626,
+      "epoch": 0.82,
+      "learning_rate": 8.950000000000001e-05,
+      "loss": 0.4944,
       "step": 200
     },
     {
-      "epoch": 0.16,
-      "eval_accuracy": 0.6746268656716418,
-      "eval_loss": 0.7115103006362915,
-      "eval_runtime": 56.506,
-      "eval_samples_per_second": 5.929,
-      "eval_steps_per_second": 1.487,
+      "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,
       "step": 200
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 9.80430528375734e-05,
-      "loss": 0.5604,
+      "epoch": 1.02,
+      "learning_rate": 8.116666666666667e-05,
+      "loss": 0.508,
       "step": 250
     },
     {
-      "epoch": 0.2,
-      "eval_accuracy": 0.6746268656716418,
-      "eval_loss": 0.700947105884552,
-      "eval_runtime": 56.4927,
-      "eval_samples_per_second": 5.93,
-      "eval_steps_per_second": 1.487,
+      "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,
       "step": 250
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 9.315068493150684e-05,
-      "loss": 0.5675,
+      "epoch": 1.23,
+      "learning_rate": 7.283333333333335e-05,
+      "loss": 0.4904,
       "step": 300
     },
     {
-      "epoch": 0.24,
-      "eval_accuracy": 0.6597014925373135,
-      "eval_loss": 0.662155032157898,
-      "eval_runtime": 56.4903,
-      "eval_samples_per_second": 5.93,
-      "eval_steps_per_second": 1.487,
+      "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,
       "step": 300
     },
     {
-      "epoch": 0.28,
-      "learning_rate": 8.825831702544032e-05,
-      "loss": 0.5725,
+      "epoch": 1.43,
+      "learning_rate": 6.450000000000001e-05,
+      "loss": 0.4694,
       "step": 350
     },
     {
-      "epoch": 0.28,
-      "eval_accuracy": 0.6567164179104478,
-      "eval_loss": 0.6746364831924438,
-      "eval_runtime": 56.482,
-      "eval_samples_per_second": 5.931,
-      "eval_steps_per_second": 1.487,
+      "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,
       "step": 350
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 8.336594911937378e-05,
-      "loss": 0.5884,
+      "epoch": 1.64,
+      "learning_rate": 5.6166666666666665e-05,
+      "loss": 0.4937,
       "step": 400
     },
     {
-      "epoch": 0.32,
-      "eval_accuracy": 0.6656716417910448,
-      "eval_loss": 0.7047849893569946,
-      "eval_runtime": 56.4873,
-      "eval_samples_per_second": 5.931,
-      "eval_steps_per_second": 1.487,
+      "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,
       "step": 400
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 7.847358121330725e-05,
-      "loss": 0.5567,
+      "epoch": 1.84,
+      "learning_rate": 4.7833333333333335e-05,
+      "loss": 0.4877,
       "step": 450
     },
     {
-      "epoch": 0.36,
+      "epoch": 1.84,
       "eval_accuracy": 0.6477611940298508,
-      "eval_loss": 0.6832321286201477,
-      "eval_runtime": 56.4814,
-      "eval_samples_per_second": 5.931,
-      "eval_steps_per_second": 1.487,
+      "eval_loss": 0.9591583609580994,
+      "eval_runtime": 56.4028,
+      "eval_samples_per_second": 5.939,
+      "eval_steps_per_second": 1.489,
       "step": 450
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 7.35812133072407e-05,
-      "loss": 0.5723,
+      "epoch": 2.05,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 0.482,
       "step": 500
     },
     {
-      "epoch": 0.4,
-      "eval_accuracy": 0.6955223880597015,
-      "eval_loss": 0.7437570691108704,
-      "eval_runtime": 56.4677,
-      "eval_samples_per_second": 5.933,
-      "eval_steps_per_second": 1.488,
+      "epoch": 2.05,
+      "eval_accuracy": 0.6716417910447762,
+      "eval_loss": 0.89241623878479,
+      "eval_runtime": 56.408,
+      "eval_samples_per_second": 5.939,
+      "eval_steps_per_second": 1.489,
       "step": 500
     }
   ],
   "logging_steps": 50,
-  "max_steps": 1247,
-  "num_train_epochs": 1,
+  "max_steps": 732,
+  "num_train_epochs": 3,
   "save_steps": 50,
-  "total_flos": 1.8605381592588288e+17,
+  "total_flos": 2.13295268918784e+17,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:87bd30bd4b21b4a0c66f1fd545858ef6e2d33b5984a8d292e79011acc3affd6b
+oid sha256:9feea516a8a208c393c324d84d047c0ab5036bbdf3fb1f3eea6a2563e9c79a6b
 size 4472