Training in progress, epoch 2

logs/events.out.tfevents.1713611453.a9446dbff3d4.7440.12

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f40b6287fbf34da48d639cae3380f68354896cc8d8a16b149626a5cb8acfe201
-size 5482
+oid sha256:077e622b1ef7a9f904e281c3128cf550a8d692a3b2093bb1b70ac0a5e022354b
+size 6212

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:40c9694ee190d60f382d769e24b26fe71862c04c4247211ccc04f65f552d99a9
+oid sha256:03dfe0952f0a6030e34124601ff79f276e2ceb48a63399634f680d64e7b92c8a
 size 17549312

run-10/checkpoint-320/config.json

@@ -27,7 +27,7 @@
   "position_embedding_type": "absolute",
   "problem_type": "single_label_classification",
   "torch_dtype": "float32",
-  "transformers_version": "4.38.2",
+  "transformers_version": "4.40.0",
   "type_vocab_size": 2,
   "use_cache": true,
   "vocab_size": 30522

run-10/checkpoint-320/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:df43e1070c8179c35300d6007c86b5179fa30a375bed414bb7493824d5589c90
+oid sha256:03dfe0952f0a6030e34124601ff79f276e2ceb48a63399634f680d64e7b92c8a
 size 17549312

run-10/checkpoint-320/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9db63500231342ee5dc8a44094ace33abc529e6af8e4c27cc3727cdc1b658e26
-size 35122746
+oid sha256:cba5ab31748daa20b30c694f336c403661966591f5dd1193478cc81419d7d3bc
+size 35123898

run-10/checkpoint-320/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a3d2fae8f804d730b1cae7bbde900a00413fe63bf39745fa2e32472d277792f0
-size 14054
+oid sha256:fce5941012578d026c76372190ed75bc3a1c59455c429a4fd246677d1bfc3594
+size 14308

run-10/checkpoint-320/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:85acdd34f6f1e7e049f0788e3eb1cd47ca53a7e3f51f2309b24a53c47c6abbb3
+oid sha256:62e046c557ffe97ab8638cd9fcf0bcee8e1793184754162b8180330c58912a7e
 size 1064

run-10/checkpoint-320/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.8131115459882583,
+  "best_metric": 0.676078431372549,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-10/checkpoint-320",
-  "epoch": 5.0,
+  "epoch": 2.0,
   "eval_steps": 500,
   "global_step": 320,
   "is_hyper_param_search": true,
@@ -10,113 +10,57 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 7.969868183135986,
-      "learning_rate": 0.0006324213595544264,
-      "loss": 0.5787,
-      "step": 64
+      "grad_norm": 2.012742280960083,
+      "learning_rate": 0.0007591954920690624,
+      "loss": 0.5869,
+      "step": 160
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.773972602739726,
-      "eval_f1": 0.8070175438596492,
-      "eval_loss": 0.48913490772247314,
-      "eval_precision": 0.7040816326530612,
-      "eval_recall": 0.9452054794520548,
-      "eval_runtime": 27.7283,
-      "eval_samples_per_second": 36.858,
-      "eval_steps_per_second": 1.154,
-      "step": 64
+      "eval_accuracy": 0.6666666666666666,
+      "eval_f1": 0.0,
+      "eval_loss": 0.5785399675369263,
+      "eval_mcc": 0.0,
+      "eval_precision": 0.0,
+      "eval_recall": 0.0,
+      "eval_runtime": 1.8697,
+      "eval_samples_per_second": 681.924,
+      "eval_steps_per_second": 21.394,
+      "step": 160
     },
     {
       "epoch": 2.0,
-      "grad_norm": 2.9250471591949463,
-      "learning_rate": 0.00047431601966581977,
-      "loss": 0.46,
-      "step": 128
-    },
-    {
-      "epoch": 2.0,
-      "eval_accuracy": 0.799412915851272,
-      "eval_f1": 0.8138056312443234,
-      "eval_loss": 0.4400319457054138,
-      "eval_precision": 0.7593220338983051,
-      "eval_recall": 0.8767123287671232,
-      "eval_runtime": 28.2626,
-      "eval_samples_per_second": 36.161,
-      "eval_steps_per_second": 1.132,
-      "step": 128
-    },
-    {
-      "epoch": 3.0,
-      "grad_norm": 3.8848695755004883,
-      "learning_rate": 0.0003162106797772132,
-      "loss": 0.4251,
-      "step": 192
-    },
-    {
-      "epoch": 3.0,
-      "eval_accuracy": 0.8111545988258317,
-      "eval_f1": 0.8170616113744076,
-      "eval_loss": 0.4429156482219696,
-      "eval_precision": 0.7922794117647058,
-      "eval_recall": 0.8434442270058709,
-      "eval_runtime": 29.3102,
-      "eval_samples_per_second": 34.868,
-      "eval_steps_per_second": 1.092,
-      "step": 192
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 6.6206135749816895,
-      "learning_rate": 0.0001581053398886066,
-      "loss": 0.4016,
-      "step": 256
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.8023483365949119,
-      "eval_f1": 0.8027343750000001,
-      "eval_loss": 0.42793938517570496,
-      "eval_precision": 0.8011695906432749,
-      "eval_recall": 0.8043052837573386,
-      "eval_runtime": 28.4384,
-      "eval_samples_per_second": 35.937,
-      "eval_steps_per_second": 1.125,
-      "step": 256
-    },
-    {
-      "epoch": 5.0,
-      "grad_norm": 2.5236122608184814,
-      "learning_rate": 0.0,
-      "loss": 0.3827,
+      "grad_norm": 2.057482957839966,
+      "learning_rate": 0.0006642960555604296,
+      "loss": 0.5715,
       "step": 320
     },
     {
-      "epoch": 5.0,
-      "eval_accuracy": 0.8131115459882583,
-      "eval_f1": 0.8242870285188593,
-      "eval_loss": 0.4034118950366974,
-      "eval_precision": 0.7777777777777778,
-      "eval_recall": 0.8767123287671232,
-      "eval_runtime": 27.9789,
-      "eval_samples_per_second": 36.528,
-      "eval_steps_per_second": 1.144,
+      "epoch": 2.0,
+      "eval_accuracy": 0.676078431372549,
+      "eval_f1": 0.4411366711772666,
+      "eval_loss": 0.5550761222839355,
+      "eval_mcc": 0.22526645932553852,
+      "eval_precision": 0.5191082802547771,
+      "eval_recall": 0.3835294117647059,
+      "eval_runtime": 1.8672,
+      "eval_samples_per_second": 682.827,
+      "eval_steps_per_second": 21.422,
       "step": 320
     }
   ],
   "logging_steps": 500,
-  "max_steps": 320,
+  "max_steps": 1440,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 9,
   "save_steps": 500,
-  "total_flos": 1178475986400.0,
-  "train_batch_size": 48,
+  "total_flos": 583510875840.0,
+  "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.9990297556295253,
-    "learning_rate": 0.000790526699443033,
-    "num_train_epochs": 5,
-    "per_device_train_batch_size": 48,
-    "temperature": 30
+    "alpha": 0.9116955099903541,
+    "learning_rate": 0.0008540949285776952,
+    "num_train_epochs": 9,
+    "temperature": 21
   }
 }

run-10/checkpoint-320/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f10318ca028e91ec5668eb2cb6ad3dd869bdfca116061dca662c8b2df479ddbc
-size 4920
+oid sha256:34acf28a009a32e591bbfa0e1392d8f5ce60c738ac5de3b2d0a889c73580dc19
+size 5048