Training in progress, epoch 2

logs/events.out.tfevents.1711285219.73e3a81c01ef.4225.14

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3fe512ac3a2f5d1d2e26bab9b3b143c189c482532ef7efd24d053bb1c83b3dd1
-size 5407
+oid sha256:4e657ceb9df26626779835d966cc10522f0888bec122b323dd546edb29c877c6
+size 6137

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:51ceb32a177137d0a3ec741c3ca5c570c990c17af4f68249b86725748d70f892
+oid sha256:340a2df6e568a766d5864c97eb83584fd8b2ddb2c4e421c26bc0523c85cbd8ac
 size 17549312

run-14/checkpoint-384/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": "4.39.1",
   "type_vocab_size": 2,
   "use_cache": true,
   "vocab_size": 30522

run-14/checkpoint-384/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2b125aea8b5ab838a483c1cc4b76fbccc277967a11bf738520b7e90b99746172
+oid sha256:340a2df6e568a766d5864c97eb83584fd8b2ddb2c4e421c26bc0523c85cbd8ac
 size 17549312

run-14/checkpoint-384/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:af268025c4e6d26b091b4083d8857658bece58c282fcff4593f457c1f0c7f456
+oid sha256:65da7f14f55ea7e17eb507a7b576b6dee4cc84f81d2d2690ce02452c957fe759
 size 35122746

run-14/checkpoint-384/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:45b1fc07ed8b121716021ec87c686f1cd9c3b89c82ea08a6a0792d47a39077c9
+oid sha256:d2558f72cda987826e5e7caf54cc6282fe335ebecbddfed7bb83f0184d1f54cc
 size 14054

run-14/checkpoint-384/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a06134b5908024e7714ec06ae17ec78fa4f915f94ad32b40348949bc8dcb767
+oid sha256:b84919ed1fee104066235977b55f934ba039e4af7ec7d16979b9d629ed72bc02
 size 1064

run-14/checkpoint-384/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.8131115459882583,
+  "best_metric": 0.7798434442270059,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-14/checkpoint-384",
-  "epoch": 4.0,
+  "epoch": 2.0,
   "eval_steps": 500,
   "global_step": 384,
   "is_hyper_param_search": true,
@@ -10,93 +10,58 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 1.392162561416626,
-      "learning_rate": 0.00018327679510936405,
-      "loss": 0.506,
-      "step": 96
+      "grad_norm": 2.0995702743530273,
+      "learning_rate": 0.00024019499178328863,
+      "loss": 0.4158,
+      "step": 192
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7945205479452054,
-      "eval_f1": 0.8138297872340425,
-      "eval_loss": 0.43322718143463135,
-      "eval_precision": 0.7439222042139384,
-      "eval_recall": 0.898238747553816,
-      "eval_runtime": 28.335,
-      "eval_samples_per_second": 36.068,
-      "eval_steps_per_second": 1.129,
-      "step": 96
-    },
-    {
-      "epoch": 2.0,
-      "grad_norm": 2.713252544403076,
-      "learning_rate": 0.00012218453007290937,
-      "loss": 0.4356,
+      "eval_accuracy": 0.7318982387475538,
+      "eval_f1": 0.7078891257995735,
+      "eval_loss": 0.38118404150009155,
+      "eval_mcc": 0.4701927470442256,
+      "eval_precision": 0.7775175644028103,
+      "eval_recall": 0.649706457925636,
+      "eval_runtime": 66.3838,
+      "eval_samples_per_second": 15.395,
+      "eval_steps_per_second": 0.482,
       "step": 192
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7974559686888454,
-      "eval_f1": 0.8169761273209548,
-      "eval_loss": 0.4169149696826935,
-      "eval_precision": 0.7451612903225806,
-      "eval_recall": 0.9041095890410958,
-      "eval_runtime": 28.1818,
-      "eval_samples_per_second": 36.265,
-      "eval_steps_per_second": 1.135,
-      "step": 192
-    },
-    {
-      "epoch": 3.0,
-      "grad_norm": 2.8020107746124268,
-      "learning_rate": 6.109226503645469e-05,
-      "loss": 0.4193,
-      "step": 288
-    },
-    {
-      "epoch": 3.0,
-      "eval_accuracy": 0.8072407045009785,
-      "eval_f1": 0.830318690783807,
-      "eval_loss": 0.4170421361923218,
-      "eval_precision": 0.7415384615384616,
-      "eval_recall": 0.9432485322896281,
-      "eval_runtime": 28.2719,
-      "eval_samples_per_second": 36.149,
-      "eval_steps_per_second": 1.132,
-      "step": 288
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 2.38786244392395,
-      "learning_rate": 0.0,
-      "loss": 0.408,
+      "grad_norm": 4.5503435134887695,
+      "learning_rate": 0.00020016249315274053,
+      "loss": 0.3871,
       "step": 384
     },
     {
-      "epoch": 4.0,
-      "eval_accuracy": 0.8131115459882583,
-      "eval_f1": 0.8311229000884174,
-      "eval_loss": 0.4037374258041382,
-      "eval_precision": 0.7580645161290323,
-      "eval_recall": 0.9197651663405088,
-      "eval_runtime": 27.777,
-      "eval_samples_per_second": 36.793,
-      "eval_steps_per_second": 1.152,
+      "epoch": 2.0,
+      "eval_accuracy": 0.7798434442270059,
+      "eval_f1": 0.7619047619047619,
+      "eval_loss": 0.3572150468826294,
+      "eval_mcc": 0.566151295376701,
+      "eval_precision": 0.8294930875576036,
+      "eval_recall": 0.7045009784735812,
+      "eval_runtime": 66.9799,
+      "eval_samples_per_second": 15.258,
+      "eval_steps_per_second": 0.478,
       "step": 384
     }
   ],
   "logging_steps": 500,
-  "max_steps": 384,
+  "max_steps": 1344,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 7,
   "save_steps": 500,
-  "total_flos": 942780789120.0,
-  "train_batch_size": 32,
+  "total_flos": 471390394560.0,
+  "train_batch_size": 16,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7512024699670536,
-    "learning_rate": 0.00024436906014581875,
-    "num_train_epochs": 4,
-    "temperature": 25
+    "alpha": 0.6115137465963567,
+    "learning_rate": 0.00028022749041383674,
+    "num_train_epochs": 7,
+    "per_device_train_batch_size": 16,
+    "temperature": 33
   }
 }

run-14/checkpoint-384/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:79a362a16e304bed4ef4106f9a75a81f97a0fe20c647127af248c0389326eed5
-size 4920
+oid sha256:0fa71636334b8b1a2e6c3a2e37b79eb3705536c6551ba1de03f195a8b0a8fb30
+size 4984