Training in progress, epoch 2

logs/events.out.tfevents.1719307811.852b1e905a9a.223.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d61a8e7f8441a98d13b2fc7109a4e918a085863ec9654b397902272fe1dedc5
-size 5429
+oid sha256:b00945b76bc6015cad282c87a8b66f8542fa67f7ac4e37499a7e4517957f74f8
+size 5963

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bb1ca79a9bf8b490a2f6783a131c6de76de46d7d9d225d392b55c2bf9992c8ad
+oid sha256:df3533a341bbd3e710d9d7de936a16f763971513d1b160bf88191e12c65b39f3
 size 17549312

run-1/checkpoint-1054/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a04314b749d4acea219b75473b7159304fd15abbb0749b74b9a72dc41ad508ee
+oid sha256:df3533a341bbd3e710d9d7de936a16f763971513d1b160bf88191e12c65b39f3
 size 17549312

run-1/checkpoint-1054/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:423e71317f9893f34e82137928e32e1c64e9bfc71f968a12029e3756d7667ebc
+oid sha256:ca9fd389e66c6e57e60204f6358fd64cb18a8d3f2f4fdb1dc1ae97e2e0ddb33a
 size 35123898

run-1/checkpoint-1054/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:02f3e3377f11bafe7b5302825b91f3a273d3263332c198caed974a11f7c0d037
+oid sha256:32816b03f4a51f12a0135d51aa95e2ea9f3d7aef81173538a39d766a38cb4af4
 size 1064

run-1/checkpoint-1054/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.8176605504587156,
+  "best_metric": 0.819954128440367,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-1054",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,41 +10,41 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 30.242748260498047,
-      "learning_rate": 0.000111895725910844,
-      "loss": 2.3991,
+      "grad_norm": 17.658132553100586,
+      "learning_rate": 7.63128750557968e-05,
+      "loss": 1.3361,
       "step": 527
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8153669724770642,
-      "eval_loss": 1.9181628227233887,
-      "eval_runtime": 2.3714,
-      "eval_samples_per_second": 367.708,
-      "eval_steps_per_second": 2.952,
+      "eval_accuracy": 0.8107798165137615,
+      "eval_loss": 1.132369041442871,
+      "eval_runtime": 2.6024,
+      "eval_samples_per_second": 335.073,
+      "eval_steps_per_second": 2.69,
       "step": 527
     },
     {
       "epoch": 2.0,
-      "grad_norm": 76.58197784423828,
-      "learning_rate": 8.95250705409598e-05,
-      "loss": 1.1199,
+      "grad_norm": 17.63370704650879,
+      "learning_rate": 3.81564375278984e-05,
+      "loss": 0.7175,
       "step": 1054
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8176605504587156,
-      "eval_loss": 1.8561229705810547,
-      "eval_runtime": 2.3966,
-      "eval_samples_per_second": 363.855,
-      "eval_steps_per_second": 2.921,
+      "eval_accuracy": 0.819954128440367,
+      "eval_loss": 1.094281554222107,
+      "eval_runtime": 2.5939,
+      "eval_samples_per_second": 336.174,
+      "eval_steps_per_second": 2.699,
       "step": 1054
     }
   ],
   "logging_steps": 500,
-  "max_steps": 3162,
+  "max_steps": 1581,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 6,
+  "num_train_epochs": 3,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -62,9 +62,9 @@
   "train_batch_size": 128,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.051381532356201576,
-    "learning_rate": 0.0001342239322193053,
-    "num_train_epochs": 6,
-    "temperature": 23
+    "alpha": 0.5918699224275719,
+    "learning_rate": 0.00011446931258369519,
+    "num_train_epochs": 3,
+    "temperature": 9
   }
 }

run-1/checkpoint-1054/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa6a7f7b16042aa90d6a482b54cde9c8a264f21ae121942f35d43fb01c1ae0eb
+oid sha256:cdb44a8bec3049f59e3c6aa052eb44c31862f011784d5f05f8553620eb8f64a0
 size 5176