Training in progress, epoch 2

logs/events.out.tfevents.1711265653.73e3a81c01ef.4225.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:29328cc7547f41743a4985f8d99faf99f508f44dea209ce91b456a7d804ef40f
-size 5407
+oid sha256:de2dbc4a46dedb66ca18a4084f06ef7d814451bb56e6f5ec8f784252448c5cd6
+size 6137

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:70b7f7f1a78e0b340b57dfae4e8a71ef78a3dbee8319a311f6096e6cf4233908
+oid sha256:4d900e4f6cb14316af78ee923303ad84861c097630cd22630ae781e3dd5558d9
 size 17549312

run-0/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-0/checkpoint-384/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1bce608ff84d676ff00ffe40bd50b97fbc7541a452729c9f63e6157c1aa613a0
+oid sha256:4d900e4f6cb14316af78ee923303ad84861c097630cd22630ae781e3dd5558d9
 size 17549312

run-0/checkpoint-384/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:39a4a9968d3befda2b3b1fc8eb232f4dd5dd4b1439f48ad4b5bd16b2994c1dd9
+oid sha256:6262f90831477c0022202b21b95260de760334c7735499127b401b2b7aad5e1f
 size 35122746

run-0/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-0/checkpoint-384/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:36f2ca0dfd9a3945022ef05581370b90827ced4c3be77e49482b876673eb94ec
+oid sha256:a5c14cda532036bd7a3de8ef7f02066ddc1a934accb6239ec8906df7f1daf52f
 size 1064

run-0/checkpoint-384/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.8003913894324853,
+  "best_metric": 0.7181996086105675,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/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": 1.6740443706512451,
-      "learning_rate": 0.0001218354408608861,
-      "loss": 0.4816,
-      "step": 96
+      "grad_norm": 1.7846055030822754,
+      "learning_rate": 4.610177667546352e-05,
+      "loss": 0.4587,
+      "step": 192
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7514677103718199,
-      "eval_f1": 0.7519531250000001,
-      "eval_loss": 0.41538161039352417,
-      "eval_precision": 0.7504873294346979,
-      "eval_recall": 0.7534246575342466,
-      "eval_runtime": 26.5175,
-      "eval_samples_per_second": 38.541,
-      "eval_steps_per_second": 1.207,
-      "step": 96
-    },
-    {
-      "epoch": 2.0,
-      "grad_norm": 1.7866544723510742,
-      "learning_rate": 9.137658064566457e-05,
-      "loss": 0.4182,
+      "eval_accuracy": 0.6976516634050881,
+      "eval_f1": 0.6419466975666281,
+      "eval_loss": 0.4008609354496002,
+      "eval_mcc": 0.41595144404027,
+      "eval_precision": 0.7869318181818182,
+      "eval_recall": 0.5420743639921722,
+      "eval_runtime": 66.5688,
+      "eval_samples_per_second": 15.353,
+      "eval_steps_per_second": 0.481,
       "step": 192
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7798434442270059,
-      "eval_f1": 0.7817652764306499,
-      "eval_loss": 0.3980446457862854,
-      "eval_precision": 0.775,
-      "eval_recall": 0.7886497064579256,
-      "eval_runtime": 25.5088,
-      "eval_samples_per_second": 40.065,
-      "eval_steps_per_second": 1.254,
-      "step": 192
-    },
-    {
-      "epoch": 3.0,
-      "grad_norm": 2.1238555908203125,
-      "learning_rate": 6.091772043044305e-05,
-      "loss": 0.4044,
-      "step": 288
-    },
-    {
-      "epoch": 3.0,
-      "eval_accuracy": 0.7984344422700587,
-      "eval_f1": 0.8205574912891985,
-      "eval_loss": 0.39847832918167114,
-      "eval_precision": 0.7394034536891679,
-      "eval_recall": 0.9217221135029354,
-      "eval_runtime": 25.6502,
-      "eval_samples_per_second": 39.844,
-      "eval_steps_per_second": 1.248,
-      "step": 288
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 2.6613142490386963,
-      "learning_rate": 3.0458860215221525e-05,
-      "loss": 0.3971,
+      "grad_norm": 6.001947402954102,
+      "learning_rate": 4.0979357044856464e-05,
+      "loss": 0.4016,
       "step": 384
     },
     {
-      "epoch": 4.0,
-      "eval_accuracy": 0.8003913894324853,
-      "eval_f1": 0.799212598425197,
-      "eval_loss": 0.38756656646728516,
-      "eval_precision": 0.803960396039604,
-      "eval_recall": 0.7945205479452054,
-      "eval_runtime": 26.1487,
-      "eval_samples_per_second": 39.084,
-      "eval_steps_per_second": 1.224,
+      "epoch": 2.0,
+      "eval_accuracy": 0.7181996086105675,
+      "eval_f1": 0.6778523489932886,
+      "eval_loss": 0.3820359408855438,
+      "eval_mcc": 0.4507700437564526,
+      "eval_precision": 0.7911227154046997,
+      "eval_recall": 0.5929549902152642,
+      "eval_runtime": 66.6005,
+      "eval_samples_per_second": 15.345,
+      "eval_steps_per_second": 0.48,
       "step": 384
     }
   ],
   "logging_steps": 500,
-  "max_steps": 480,
+  "max_steps": 1920,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 10,
   "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": 0.679174768290245,
-    "learning_rate": 0.00015229430107610762,
-    "num_train_epochs": 5,
-    "temperature": 27
+    "alpha": 0.6226480553446316,
+    "learning_rate": 5.122419630607058e-05,
+    "num_train_epochs": 10,
+    "per_device_train_batch_size": 16,
+    "temperature": 20
   }
 }

run-0/checkpoint-384/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1873f3a503b2d5b3f7200baa33f737ba864dd65edb3834d85c5b8e40b6b72f07
-size 4920
+oid sha256:dbca9cde0b156ba25d74ee6bc6ea4c2fa160afa3570e75ccc73231ac4a159c49
+size 4984