Training in progress, epoch 2

logs/events.out.tfevents.1709745084.1dcff7455174.7719.15

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dc2d079a64417d94df547283cfee9b20b22c0a65537ff4a7a23d46b0833e9b1f
-size 5316
+oid sha256:2b6cec3f47bf1f33878b31ad23e873cb04f69a65667d6de9924a05d280fdd389
+size 5999

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f68fb0faebee8cb7dc45d058363b1fd5ab13a67c70abd1f7c30b1ac95a145c76
+oid sha256:3de6660b415820259563363fcfe531f83c935a94cce18bbcfbb2de5b473f6be3
 size 17549312

run-15/checkpoint-192/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e51c6bffd5e80d5db594fbb33a358a6a8f49d1861e0d5cd8ab89b4941f4c1d18
+oid sha256:3de6660b415820259563363fcfe531f83c935a94cce18bbcfbb2de5b473f6be3
 size 17549312

run-15/checkpoint-192/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:58cf9e3ae69003d59abe4aa91c0e2c079ec90980d90b2d6c208e6bd3ef06eed0
+oid sha256:ea43d6f75916e25d40ef3e949954b627eb9c168a44eac89dbdb756c55220cdee
 size 35122746

run-15/checkpoint-192/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6cae425c8fb6e43e1c23cd81e499e76b6567e8693dfdeaf295bdedf3136ae373
+oid sha256:5ae96168c84278fc0fe1f596c71853c3d53ac948eeccd69bb2bfcfaac2b74b0c
 size 1064

run-15/checkpoint-192/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7984344422700587,
+  "best_metric": 0.8189823874755382,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-15/checkpoint-192",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,55 +10,55 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 1.150837779045105,
-      "learning_rate": 0.00014120530022151875,
-      "loss": 0.5007,
+      "grad_norm": 2.341264247894287,
+      "learning_rate": 0.0003257774454972516,
+      "loss": 0.5225,
       "step": 96
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7681017612524462,
-      "eval_f1": 0.7889581478183438,
-      "eval_loss": 0.4356862008571625,
-      "eval_precision": 0.7238562091503268,
-      "eval_recall": 0.8669275929549902,
-      "eval_runtime": 28.2732,
-      "eval_samples_per_second": 36.147,
-      "eval_steps_per_second": 1.132,
+      "eval_accuracy": 0.7906066536203522,
+      "eval_f1": 0.8078994614003591,
+      "eval_loss": 0.4404396712779999,
+      "eval_precision": 0.746268656716418,
+      "eval_recall": 0.8806262230919765,
+      "eval_runtime": 24.9228,
+      "eval_samples_per_second": 41.007,
+      "eval_steps_per_second": 1.284,
       "step": 96
     },
     {
       "epoch": 2.0,
-      "grad_norm": 2.0905497074127197,
-      "learning_rate": 9.413686681434584e-05,
-      "loss": 0.4302,
+      "grad_norm": 2.7467563152313232,
+      "learning_rate": 0.00028957995155311255,
+      "loss": 0.4443,
       "step": 192
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7984344422700587,
-      "eval_f1": 0.81993006993007,
-      "eval_loss": 0.4125403165817261,
-      "eval_precision": 0.740916271721959,
-      "eval_recall": 0.9178082191780822,
-      "eval_runtime": 29.852,
-      "eval_samples_per_second": 34.236,
-      "eval_steps_per_second": 1.072,
+      "eval_accuracy": 0.8189823874755382,
+      "eval_f1": 0.8322756119673618,
+      "eval_loss": 0.40899789333343506,
+      "eval_precision": 0.7753378378378378,
+      "eval_recall": 0.898238747553816,
+      "eval_runtime": 24.9291,
+      "eval_samples_per_second": 40.996,
+      "eval_steps_per_second": 1.284,
       "step": 192
     }
   ],
   "logging_steps": 500,
-  "max_steps": 384,
+  "max_steps": 960,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 10,
   "save_steps": 500,
   "total_flos": 471390394560.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7235742113106859,
-    "learning_rate": 0.00018827373362869168,
-    "num_train_epochs": 4,
-    "temperature": 25
+    "alpha": 0.7955319016044536,
+    "learning_rate": 0.00036197493944139066,
+    "num_train_epochs": 10,
+    "temperature": 8
   }
 }

run-15/checkpoint-192/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b4455ec233a06250ca78d744f669f3caf07ff1e989cb0faa606b8567c0ee6a96
+oid sha256:e3d9bd9e7f16ac97ca0e7e00ad365e1b9f7804af850a7b1faa4910402684f951
 size 4920