Training in progress, epoch 1

config.json

@@ -15,8 +15,8 @@
   "initializer_range": 0.02,
   "intermediate_size": 512,
   "label2id": {
-    "negative": 0,
-    "positive": 1
+    "negative": "0",
+    "positive": "1"
   },
   "layer_norm_eps": 1e-12,
   "max_position_embeddings": 512,

logs/events.out.tfevents.1711296858.8cc2eef2edb7.2942.2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2639c1c7eed6ff11aac2694a7d4bb784db95973d8617cf620672b55879543225
-size 6137
+oid sha256:eec5f8a9d179cd75412d3f476893b961e415378f7736f0264bbfd87b7d5a61d3
+size 11601

logs/events.out.tfevents.1711297031.8cc2eef2edb7.2942.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5671c871dec21b814e238027bc460e37eeca98777d4d78926a461e91685a1999
+size 5406

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:21c84e6f83ea178d5ec19c03e2313d64e2ee2968539fb044cd51b74d740b3c4f
+oid sha256:8535679c03461b298018d74a95c8d714c32a4f38defdc9aadd1837a148497803
 size 17549312

run-2/checkpoint-2673/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-2/checkpoint-2673/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:24d3b3773a2cb5eb4fe5d0510281ec4afc468d94f153ea1c4e42557acc72d031
+oid sha256:02cb61417b7bc4627bef9aead38201284ebffcb7218d1cd7f8e50100028fec79
 size 17549312

run-2/checkpoint-2673/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aaaf77f832395c735d2ca2829839fa8b3de3e358a4b7cfb575a266ac248c48c0
-size 35122746
+oid sha256:01ae6ef254146e91c7ec1fd3d28b8863e057d4dc9c23f3165352d58d53bb29ef
+size 35123898

run-2/checkpoint-2673/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cb49527a73de8c97b0805cc04f8d3e06cb2a963ae0f111d98648316d291a7702
-size 14054
+oid sha256:7a4a1b62f050c820b6eb360863e5fa90a75f0572b605f7ec9339d28261a796fd
+size 14308

run-2/checkpoint-2673/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d88a949c40227921ebbc201965ce805890fc5d6557c00c377fb1db010a9e5d7
+oid sha256:ecc18bc8ae20249ae705a82acbac424d3d36024accd1565a75c4986b5d70555c
 size 1064

run-2/checkpoint-2673/tokenizer.json

@@ -2,7 +2,7 @@
   "version": "1.0",
   "truncation": {
     "direction": "Right",
-    "max_length": 33,
+    "max_length": 31,
     "strategy": "LongestFirst",
     "stride": 0
   },

run-2/checkpoint-2673/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.5207920792079208,
+  "best_metric": 0.5128712871287129,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-594",
   "epoch": 9.0,
   "eval_steps": 500,
@@ -10,146 +10,182 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 0.46941760182380676,
-      "learning_rate": 1.0066413527942762e-05,
-      "loss": 0.2155,
+      "grad_norm": 0.6258772015571594,
+      "learning_rate": 1.171336966673164e-05,
+      "loss": 0.303,
       "step": 297
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.497029702970297,
-      "eval_loss": 0.18417517840862274,
-      "eval_runtime": 14.6034,
-      "eval_samples_per_second": 34.581,
-      "eval_steps_per_second": 1.096,
+      "eval_accuracy": 0.4910891089108911,
+      "eval_f1": 0.0,
+      "eval_loss": 0.28316354751586914,
+      "eval_mcc": -0.09980217586956908,
+      "eval_precision": 0.0,
+      "eval_recall": 0.0,
+      "eval_runtime": 0.9281,
+      "eval_samples_per_second": 544.113,
+      "eval_steps_per_second": 17.239,
       "step": 297
     },
     {
       "epoch": 2.0,
-      "grad_norm": 0.3071568012237549,
-      "learning_rate": 8.808111836949918e-06,
-      "loss": 0.1831,
+      "grad_norm": 0.44756296277046204,
+      "learning_rate": 1.0249198458390185e-05,
+      "loss": 0.2807,
       "step": 594
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5207920792079208,
-      "eval_loss": 0.17807520925998688,
-      "eval_runtime": 14.6044,
-      "eval_samples_per_second": 34.579,
-      "eval_steps_per_second": 1.096,
+      "eval_accuracy": 0.5128712871287129,
+      "eval_f1": 0.13986013986013984,
+      "eval_loss": 0.2766987085342407,
+      "eval_mcc": 0.047945551909639166,
+      "eval_precision": 0.5882352941176471,
+      "eval_recall": 0.07936507936507936,
+      "eval_runtime": 0.9353,
+      "eval_samples_per_second": 539.951,
+      "eval_steps_per_second": 17.107,
       "step": 594
     },
     {
       "epoch": 3.0,
-      "grad_norm": 0.35216400027275085,
-      "learning_rate": 7.549810145957073e-06,
-      "loss": 0.1804,
+      "grad_norm": 0.431320458650589,
+      "learning_rate": 8.785027250048729e-06,
+      "loss": 0.2776,
       "step": 891
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.5207920792079208,
-      "eval_loss": 0.17647753655910492,
-      "eval_runtime": 15.4786,
-      "eval_samples_per_second": 32.626,
-      "eval_steps_per_second": 1.034,
+      "eval_accuracy": 0.5128712871287129,
+      "eval_f1": 0.14583333333333331,
+      "eval_loss": 0.27507784962654114,
+      "eval_mcc": 0.046724423554504804,
+      "eval_precision": 0.5833333333333334,
+      "eval_recall": 0.08333333333333333,
+      "eval_runtime": 0.9247,
+      "eval_samples_per_second": 546.126,
+      "eval_steps_per_second": 17.303,
       "step": 891
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.26479315757751465,
-      "learning_rate": 6.291508454964228e-06,
-      "loss": 0.179,
+      "grad_norm": 0.3838063180446625,
+      "learning_rate": 7.320856041707276e-06,
+      "loss": 0.2774,
       "step": 1188
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.5168316831683168,
-      "eval_loss": 0.1757788509130478,
-      "eval_runtime": 13.9632,
-      "eval_samples_per_second": 36.166,
-      "eval_steps_per_second": 1.146,
+      "eval_accuracy": 0.5128712871287129,
+      "eval_f1": 0.14583333333333331,
+      "eval_loss": 0.273956835269928,
+      "eval_mcc": 0.046724423554504804,
+      "eval_precision": 0.5833333333333334,
+      "eval_recall": 0.08333333333333333,
+      "eval_runtime": 0.9281,
+      "eval_samples_per_second": 544.106,
+      "eval_steps_per_second": 17.239,
       "step": 1188
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.33599618077278137,
-      "learning_rate": 5.033206763971381e-06,
-      "loss": 0.1781,
+      "grad_norm": 0.5975275039672852,
+      "learning_rate": 5.85668483336582e-06,
+      "loss": 0.2766,
       "step": 1485
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.5168316831683168,
-      "eval_loss": 0.17515264451503754,
-      "eval_runtime": 14.2891,
-      "eval_samples_per_second": 35.342,
-      "eval_steps_per_second": 1.12,
+      "eval_accuracy": 0.5108910891089109,
+      "eval_f1": 0.13937282229965156,
+      "eval_loss": 0.27307695150375366,
+      "eval_mcc": 0.039524509362779174,
+      "eval_precision": 0.5714285714285714,
+      "eval_recall": 0.07936507936507936,
+      "eval_runtime": 0.9415,
+      "eval_samples_per_second": 536.374,
+      "eval_steps_per_second": 16.994,
       "step": 1485
     },
     {
       "epoch": 6.0,
-      "grad_norm": 0.21752449870109558,
-      "learning_rate": 3.7749050729785363e-06,
-      "loss": 0.1781,
+      "grad_norm": 0.4973074495792389,
+      "learning_rate": 4.3925136250243645e-06,
+      "loss": 0.2759,
       "step": 1782
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.5168316831683168,
-      "eval_loss": 0.17485138773918152,
-      "eval_runtime": 13.9491,
-      "eval_samples_per_second": 36.203,
-      "eval_steps_per_second": 1.147,
+      "eval_accuracy": 0.5069306930693069,
+      "eval_f1": 0.1384083044982699,
+      "eval_loss": 0.27315467596054077,
+      "eval_mcc": 0.023354867176075732,
+      "eval_precision": 0.5405405405405406,
+      "eval_recall": 0.07936507936507936,
+      "eval_runtime": 0.9268,
+      "eval_samples_per_second": 544.884,
+      "eval_steps_per_second": 17.264,
       "step": 1782
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.2491326630115509,
-      "learning_rate": 2.5166033819856906e-06,
-      "loss": 0.1777,
+      "grad_norm": 0.440789133310318,
+      "learning_rate": 2.92834241668291e-06,
+      "loss": 0.2755,
       "step": 2079
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.5168316831683168,
-      "eval_loss": 0.17470771074295044,
-      "eval_runtime": 14.4075,
-      "eval_samples_per_second": 35.051,
-      "eval_steps_per_second": 1.111,
+      "eval_accuracy": 0.5108910891089109,
+      "eval_f1": 0.13937282229965156,
+      "eval_loss": 0.27257436513900757,
+      "eval_mcc": 0.039524509362779174,
+      "eval_precision": 0.5714285714285714,
+      "eval_recall": 0.07936507936507936,
+      "eval_runtime": 0.9289,
+      "eval_samples_per_second": 543.634,
+      "eval_steps_per_second": 17.224,
       "step": 2079
     },
     {
       "epoch": 8.0,
-      "grad_norm": 0.5971789956092834,
-      "learning_rate": 1.2583016909928453e-06,
-      "loss": 0.1776,
+      "grad_norm": 0.6993103623390198,
+      "learning_rate": 1.464171208341455e-06,
+      "loss": 0.2751,
       "step": 2376
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.5168316831683168,
-      "eval_loss": 0.17442195117473602,
-      "eval_runtime": 14.8484,
-      "eval_samples_per_second": 34.011,
-      "eval_steps_per_second": 1.078,
+      "eval_accuracy": 0.5089108910891089,
+      "eval_f1": 0.1267605633802817,
+      "eval_loss": 0.27230069041252136,
+      "eval_mcc": 0.03302792352019969,
+      "eval_precision": 0.5625,
+      "eval_recall": 0.07142857142857142,
+      "eval_runtime": 0.9319,
+      "eval_samples_per_second": 541.932,
+      "eval_steps_per_second": 17.17,
       "step": 2376
     },
     {
       "epoch": 9.0,
-      "grad_norm": 0.2817859947681427,
+      "grad_norm": 0.5155681371688843,
       "learning_rate": 0.0,
-      "loss": 0.1773,
+      "loss": 0.2752,
       "step": 2673
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.5168316831683168,
-      "eval_loss": 0.17444071173667908,
-      "eval_runtime": 13.7553,
-      "eval_samples_per_second": 36.713,
-      "eval_steps_per_second": 1.163,
+      "eval_accuracy": 0.5069306930693069,
+      "eval_f1": 0.1263157894736842,
+      "eval_loss": 0.2723012864589691,
+      "eval_mcc": 0.024561362814430283,
+      "eval_precision": 0.5454545454545454,
+      "eval_recall": 0.07142857142857142,
+      "eval_runtime": 0.9301,
+      "eval_samples_per_second": 542.98,
+      "eval_steps_per_second": 17.203,
       "step": 2673
     }
   ],
@@ -158,13 +194,13 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9,
   "save_steps": 500,
-  "total_flos": 7000589548260.0,
+  "total_flos": 6576311393820.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.20650329892275032,
-    "learning_rate": 1.1324715218935609e-05,
+    "alpha": 0.33782488262757904,
+    "learning_rate": 1.3177540875073095e-05,
     "num_train_epochs": 9,
-    "temperature": 10
+    "temperature": 26
   }
 }

run-2/checkpoint-2673/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:53af5913dd7a7ac18c4af82b6aa52534a7091d8ef8c41f1056aa7fa989430b5d
-size 4984
+oid sha256:28b3023982056159c033a3a1a25c2d159010d9ff247770a44c7d7d6d0e3bba5b
+size 4920

run-3/checkpoint-297/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-3/checkpoint-297/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76220d7fcfb19b8bac2ae99022cae216a90b830bf00b3012d8b613a63d0bf947
+oid sha256:8535679c03461b298018d74a95c8d714c32a4f38defdc9aadd1837a148497803
 size 17549312

run-3/checkpoint-297/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2774d92a34e603acd644175167e2e31366f8af947ecea6c85f0689d20f0ae0b8
-size 35122746
+oid sha256:60be5d5c92b1bb86ec6b7a46b5ac62e4e96cd5078a9b61abbbb3234637c75f50
+size 35123898

run-3/checkpoint-297/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f414017d19e8a66d09e6a16c0bca909eff6c9e5541f54da3f0dba2607378e04d
-size 14054
+oid sha256:a5f5e396c574b9b2ecc40d3f9e87d628a8f4be69caef81905a47fcf8895067e6
+size 14308

run-3/checkpoint-297/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7dc0e11b8c40cbed76dc3148c9dd20b9b6423efc82e18995f762ffefe8f96c5f
+oid sha256:9270fba2490052dbb63d40feaa1bf95c7d144c1550805c5447392f630cb53251
 size 1064

run-3/checkpoint-297/tokenizer.json

@@ -2,7 +2,7 @@
   "version": "1.0",
   "truncation": {
     "direction": "Right",
-    "max_length": 33,
+    "max_length": 31,
     "strategy": "LongestFirst",
     "stride": 0
   },

run-3/checkpoint-297/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.5128712871287129,
+  "best_metric": 0.5168316831683168,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-3/checkpoint-297",
   "epoch": 1.0,
   "eval_steps": 500,
@@ -10,18 +10,22 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 0.9984288811683655,
-      "learning_rate": 9.995670829688553e-05,
-      "loss": 0.5706,
+      "grad_norm": 0.8371890187263489,
+      "learning_rate": 0.000508582557644255,
+      "loss": 0.309,
       "step": 297
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5128712871287129,
-      "eval_loss": 0.5624967217445374,
-      "eval_runtime": 13.6049,
-      "eval_samples_per_second": 37.119,
-      "eval_steps_per_second": 1.176,
+      "eval_accuracy": 0.5168316831683168,
+      "eval_f1": 0.1643835616438356,
+      "eval_loss": 0.30045250058174133,
+      "eval_mcc": 0.059239742807176775,
+      "eval_precision": 0.6,
+      "eval_recall": 0.09523809523809523,
+      "eval_runtime": 0.9336,
+      "eval_samples_per_second": 540.929,
+      "eval_steps_per_second": 17.138,
       "step": 297
     }
   ],
@@ -30,13 +34,13 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
-  "total_flos": 777843283140.0,
+  "total_flos": 730701265980.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7785816803005383,
-    "learning_rate": 0.0001499350624453283,
+    "alpha": 0.37059488996882817,
+    "learning_rate": 0.0007628738364663827,
     "num_train_epochs": 3,
-    "temperature": 16
+    "temperature": 5
   }
 }

run-3/checkpoint-297/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9decda4f5411e0ed12b24ed68c82ec0237bdb8018a369738f83b0ca5fb77d628
-size 4984
+oid sha256:c77ece67af99569b700b80b7c3d227a73e33d56768e0354e21fbf558f6ec676a
+size 4920

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f62d580257060cdaf08ca2f6bebff447fe36fd6a4ed09c360f3a936d71824ec
+oid sha256:c77ece67af99569b700b80b7c3d227a73e33d56768e0354e21fbf558f6ec676a
 size 4920