Training in progress, epoch 6

logs/events.out.tfevents.1719310313.852b1e905a9a.223.5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d753cd9c975379dd90b5d356886d24298df7ec6c21398a39b744289b5bd80a24
-size 7565
+oid sha256:6b7c27678c96b55268ae20101a3dcdcabed72a7e35b5314e74a49c2627c96740
+size 8453

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:97d865cd86bbd82a83507568d3aedd623e054710a155c7e8e3cddf0ba2b7c970
+oid sha256:1d47eed6b6c10b89046c5579792754bc1f6f31fc160278e47e2d473bc77094d3
 size 17549312

run-0/checkpoint-3162/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e66d9522c7e5449c5f8219463267e60b5ec3f1017664ef4e316e07a78f5f39e
+oid sha256:1d47eed6b6c10b89046c5579792754bc1f6f31fc160278e47e2d473bc77094d3
 size 17549312

run-0/checkpoint-3162/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:83703cd4bf2a473fb393d524ea8bded4014b2b81fcc50abc39cd466f11e50108
+oid sha256:d12374f16343dd9854da47ea59df55bcfc6c07f511e63ee993f70c9343f14ab5
 size 35123898

run-0/checkpoint-3162/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0eb013781aa27303ed3472ad10f082332665efcab61c356458c316242fc903cb
+oid sha256:4be833998b5d5f182cff6fbb7697231be6c352b034f0ce1e3cb32add5329f3a2
 size 1064

run-0/checkpoint-3162/trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.7947247706422018,
-  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-3162",
+  "best_metric": 0.801605504587156,
+  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-2108",
   "epoch": 6.0,
   "eval_steps": 500,
   "global_step": 3162,
@@ -10,105 +10,105 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 5.95961856842041,
-      "learning_rate": 9.55389368279823e-06,
-      "loss": 1.5369,
+      "grad_norm": 12.377176284790039,
+      "learning_rate": 1.5980854192157344e-05,
+      "loss": 3.0218,
       "step": 527
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7339449541284404,
-      "eval_loss": 1.2773902416229248,
-      "eval_runtime": 2.5957,
-      "eval_samples_per_second": 335.939,
-      "eval_steps_per_second": 2.697,
+      "eval_accuracy": 0.7672018348623854,
+      "eval_loss": 2.191861391067505,
+      "eval_runtime": 2.5906,
+      "eval_samples_per_second": 336.603,
+      "eval_steps_per_second": 2.702,
       "step": 527
     },
     {
       "epoch": 2.0,
-      "grad_norm": 15.252978324890137,
-      "learning_rate": 8.492349940265094e-06,
-      "loss": 1.2159,
+      "grad_norm": Infinity,
+      "learning_rate": 1.2796813034327135e-05,
+      "loss": 2.0797,
       "step": 1054
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7786697247706422,
-      "eval_loss": 1.022659182548523,
-      "eval_runtime": 2.5741,
-      "eval_samples_per_second": 338.753,
-      "eval_steps_per_second": 2.719,
+      "eval_accuracy": 0.7809633027522935,
+      "eval_loss": 1.8230279684066772,
+      "eval_runtime": 2.5719,
+      "eval_samples_per_second": 339.043,
+      "eval_steps_per_second": 2.722,
       "step": 1054
     },
     {
       "epoch": 3.0,
-      "grad_norm": 18.01114845275879,
-      "learning_rate": 7.430806197731956e-06,
-      "loss": 1.0132,
+      "grad_norm": 40.96967315673828,
+      "learning_rate": 9.600642195895665e-06,
+      "loss": 1.697,
       "step": 1581
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.7844036697247706,
-      "eval_loss": 0.9622328281402588,
-      "eval_runtime": 2.5865,
-      "eval_samples_per_second": 337.137,
-      "eval_steps_per_second": 2.706,
+      "eval_accuracy": 0.786697247706422,
+      "eval_loss": 1.7212010622024536,
+      "eval_runtime": 2.549,
+      "eval_samples_per_second": 342.099,
+      "eval_steps_per_second": 2.746,
       "step": 1581
     },
     {
       "epoch": 4.0,
-      "grad_norm": 15.912079811096191,
-      "learning_rate": 6.371276769700781e-06,
-      "loss": 0.9206,
+      "grad_norm": 35.09682846069336,
+      "learning_rate": 6.4044713574641965e-06,
+      "loss": 1.5166,
       "step": 2108
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.7901376146788991,
-      "eval_loss": 0.9278557896614075,
-      "eval_runtime": 2.5808,
-      "eval_samples_per_second": 337.878,
-      "eval_steps_per_second": 2.712,
+      "eval_accuracy": 0.801605504587156,
+      "eval_loss": 1.6234298944473267,
+      "eval_runtime": 2.5722,
+      "eval_samples_per_second": 339.013,
+      "eval_steps_per_second": 2.721,
       "step": 2108
     },
     {
       "epoch": 5.0,
-      "grad_norm": 10.149016380310059,
-      "learning_rate": 5.3097330271676446e-06,
-      "loss": 0.8564,
+      "grad_norm": 10.431685447692871,
+      "learning_rate": 3.208300519032727e-06,
+      "loss": 1.4053,
       "step": 2635
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.7901376146788991,
-      "eval_loss": 0.9079629182815552,
-      "eval_runtime": 2.5658,
-      "eval_samples_per_second": 339.852,
-      "eval_steps_per_second": 2.728,
+      "eval_accuracy": 0.7993119266055045,
+      "eval_loss": 1.5793094635009766,
+      "eval_runtime": 2.5626,
+      "eval_samples_per_second": 340.279,
+      "eval_steps_per_second": 2.732,
       "step": 2635
     },
     {
       "epoch": 6.0,
-      "grad_norm": 16.081409454345703,
-      "learning_rate": 4.248189284634507e-06,
-      "loss": 0.8139,
+      "grad_norm": 31.159494400024414,
+      "learning_rate": 1.2129680601257947e-08,
+      "loss": 1.3567,
       "step": 3162
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.7947247706422018,
-      "eval_loss": 0.8956273794174194,
-      "eval_runtime": 2.6131,
-      "eval_samples_per_second": 333.701,
-      "eval_steps_per_second": 2.679,
+      "eval_accuracy": 0.8004587155963303,
+      "eval_loss": 1.5706629753112793,
+      "eval_runtime": 2.5683,
+      "eval_samples_per_second": 339.518,
+      "eval_steps_per_second": 2.725,
       "step": 3162
     }
   ],
   "logging_steps": 500,
-  "max_steps": 5270,
+  "max_steps": 3162,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 6,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -117,7 +117,7 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
@@ -126,9 +126,9 @@
   "train_batch_size": 128,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.27608826195592573,
-    "learning_rate": 1.0615437425331367e-05,
-    "num_train_epochs": 10,
-    "temperature": 2
+    "alpha": 0.275210688972374,
+    "learning_rate": 1.9177025030588814e-05,
+    "num_train_epochs": 6,
+    "temperature": 6
   }
 }

run-0/checkpoint-3162/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8b6a60f7b85b38fa45cddf1a417ee51250fe5822237403416bf2406ff2cdb84
+oid sha256:15e7b4feae857373f91378a3d0efc15d0ec396bfba71e74c965086843aa6acf4
 size 5176