Training in progress, epoch 1

logs/events.out.tfevents.1709703029.adc675a344d5.5455.2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:88417bf6aa67a14edd1329118b7ae725f33b884f1e1d83a6a5c7812c33e7e3c0
-size 5997
+oid sha256:b984b82b2d2027d9691d6a7198899d4076333473a0276818f9f117e79fe8cc72
+size 8400

logs/events.out.tfevents.1709703603.adc675a344d5.5455.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ae3c3b08ef1769428ea140509e917722a81586905f1c56590d82048c762d96d
+size 5315

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:564f1f46932544fd89608b22974dc27996ca6dfb380e01ea2b6299c6c592a2bf
+oid sha256:f72e368a0735e0614c97feb6fba15c5d298f68104e7128cf582de5807e255027
 size 17549312

run-1/checkpoint-384/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7189ccff9386ad17f3ff71b43ae8061f4fea6bd66b1763cfed6af3fbd59e1f22
+oid sha256:58db24d6a1e94d489cf4b73b5c63cab45aa03d29d732a9d54d6f4f46aef7da02
 size 17549312

run-1/checkpoint-384/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:afb0b996b74d5df5943878af57f76dd98df2dc9b5ccf5bbe1840dd4b0ed56b28
+oid sha256:a4d03bf37ea38e30d96312f5718b7d70295e30f2d32d7091c3106fe205c75b2d
 size 35122746

run-1/checkpoint-384/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3be5f1501640a7635c7c786c03440cb1f690d52e629e5cdf40292a361f5a1028
+oid sha256:9d4b63168300ff1c1b6fb1c3579fe55a55c3b0dedd2cdac5f6ca12f671c781e4
 size 1064

run-1/checkpoint-384/trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.5,
-  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-96",
+  "best_metric": 0.8140900195694716,
+  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-288",
   "epoch": 4.0,
   "eval_steps": 500,
   "global_step": 384,
@@ -10,93 +10,93 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 0.4667714238166809,
-      "learning_rate": 7.759069356217338e-06,
-      "loss": 0.317,
+      "grad_norm": 2.6617980003356934,
+      "learning_rate": 0.0006635023570917325,
+      "loss": 0.5037,
       "step": 96
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.5,
-      "eval_f1": 0.0,
-      "eval_loss": 0.26975956559181213,
-      "eval_precision": 0.0,
-      "eval_recall": 0.0,
-      "eval_runtime": 28.3197,
-      "eval_samples_per_second": 36.088,
-      "eval_steps_per_second": 1.13,
+      "eval_accuracy": 0.7661448140900196,
+      "eval_f1": 0.792714657415438,
+      "eval_loss": 0.4430159628391266,
+      "eval_precision": 0.7118380062305296,
+      "eval_recall": 0.8943248532289628,
+      "eval_runtime": 28.6854,
+      "eval_samples_per_second": 35.628,
+      "eval_steps_per_second": 1.116,
       "step": 96
     },
     {
       "epoch": 2.0,
-      "grad_norm": 0.5722590088844299,
-      "learning_rate": 5.1727129041448925e-06,
-      "loss": 0.267,
+      "grad_norm": 1.12320077419281,
+      "learning_rate": 0.0004976267678187994,
+      "loss": 0.4391,
       "step": 192
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5,
-      "eval_f1": 0.0,
-      "eval_loss": 0.2646586298942566,
-      "eval_precision": 0.0,
-      "eval_recall": 0.0,
-      "eval_runtime": 29.2799,
-      "eval_samples_per_second": 34.905,
-      "eval_steps_per_second": 1.093,
+      "eval_accuracy": 0.8082191780821918,
+      "eval_f1": 0.8227848101265822,
+      "eval_loss": 0.41584911942481995,
+      "eval_precision": 0.7647058823529411,
+      "eval_recall": 0.8904109589041096,
+      "eval_runtime": 28.5215,
+      "eval_samples_per_second": 35.833,
+      "eval_steps_per_second": 1.122,
       "step": 192
     },
     {
       "epoch": 3.0,
-      "grad_norm": 0.38027867674827576,
-      "learning_rate": 2.5863564520724462e-06,
-      "loss": 0.2634,
+      "grad_norm": 2.946756362915039,
+      "learning_rate": 0.00033175117854586627,
+      "loss": 0.41,
       "step": 288
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.5,
-      "eval_f1": 0.0,
-      "eval_loss": 0.26031577587127686,
-      "eval_precision": 0.0,
-      "eval_recall": 0.0,
-      "eval_runtime": 28.4418,
-      "eval_samples_per_second": 35.933,
-      "eval_steps_per_second": 1.125,
+      "eval_accuracy": 0.8140900195694716,
+      "eval_f1": 0.8260073260073261,
+      "eval_loss": 0.43016910552978516,
+      "eval_precision": 0.7762478485370051,
+      "eval_recall": 0.8825831702544031,
+      "eval_runtime": 28.1244,
+      "eval_samples_per_second": 36.339,
+      "eval_steps_per_second": 1.138,
       "step": 288
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.5131752490997314,
-      "learning_rate": 0.0,
-      "loss": 0.2603,
+      "grad_norm": 3.868119478225708,
+      "learning_rate": 0.00016587558927293314,
+      "loss": 0.4041,
       "step": 384
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.5,
-      "eval_f1": 0.0,
-      "eval_loss": 0.2587716281414032,
-      "eval_precision": 0.0,
-      "eval_recall": 0.0,
-      "eval_runtime": 28.5668,
-      "eval_samples_per_second": 35.776,
-      "eval_steps_per_second": 1.12,
+      "eval_accuracy": 0.8033268101761253,
+      "eval_f1": 0.802747791952895,
+      "eval_loss": 0.4142443835735321,
+      "eval_precision": 0.8051181102362205,
+      "eval_recall": 0.8003913894324853,
+      "eval_runtime": 29.1758,
+      "eval_samples_per_second": 35.029,
+      "eval_steps_per_second": 1.097,
       "step": 384
     }
   ],
   "logging_steps": 500,
-  "max_steps": 384,
+  "max_steps": 480,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "total_flos": 942780789120.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.2995113868248127,
-    "learning_rate": 1.0345425808289785e-05,
-    "num_train_epochs": 4,
-    "temperature": 18
+    "alpha": 0.7818460169984094,
+    "learning_rate": 0.0008293779463646656,
+    "num_train_epochs": 5,
+    "temperature": 3
   }
 }

run-1/checkpoint-384/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f7a7027861ec8f0a120b010edf2e889059268bbeac3a62f529a79661a3cf64f
+oid sha256:1ab2cfe4c2ab241510bc23e9ec5727876e620ab477a4e5ad72a937b25e09df23
 size 4920

run-1/checkpoint-480/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:847df229a7ca0cb4095cbd280d7c3c672c7e4ee9114fc1bc359491abd2ab3ffb
+oid sha256:7c17b3aea0a20f269c52c31bcda653e649c0a415fe70d6bb7cfb066e7b37274c
 size 17549312

run-1/checkpoint-480/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4f70ab8c31ca18dfb957c1ba52f19a90faf43d594db85a5dde58bb6f884e93a5
+oid sha256:3ef1b604fb2af4ebef635a814702213be5553c77dc1e1ed97f5f47c3195134ec
 size 35122746

run-1/checkpoint-480/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9df0e59ca0915eb128e8de1597a625689a2ef7a8993192bedb8693c0c8a3fdb2
+oid sha256:71b7b43a8fcd1d4e4ca7f947e4f0461e06c448069a6caeee0e0a975527a03251
 size 1064

run-1/checkpoint-480/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.8277886497064579,
+  "best_metric": 0.8238747553816047,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-480",
   "epoch": 5.0,
   "eval_steps": 500,
@@ -10,97 +10,97 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 1.4932310581207275,
-      "learning_rate": 0.00022974033046670533,
-      "loss": 0.4667,
+      "grad_norm": 2.6617980003356934,
+      "learning_rate": 0.0006635023570917325,
+      "loss": 0.5037,
       "step": 96
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7651663405088063,
-      "eval_f1": 0.7701149425287357,
-      "eval_loss": 0.4088786244392395,
-      "eval_precision": 0.7542213883677298,
-      "eval_recall": 0.786692759295499,
-      "eval_runtime": 30.971,
-      "eval_samples_per_second": 32.999,
-      "eval_steps_per_second": 1.033,
+      "eval_accuracy": 0.7661448140900196,
+      "eval_f1": 0.792714657415438,
+      "eval_loss": 0.4430159628391266,
+      "eval_precision": 0.7118380062305296,
+      "eval_recall": 0.8943248532289628,
+      "eval_runtime": 28.6854,
+      "eval_samples_per_second": 35.628,
+      "eval_steps_per_second": 1.116,
       "step": 96
     },
     {
       "epoch": 2.0,
-      "grad_norm": 2.3421308994293213,
-      "learning_rate": 0.00017230524785002898,
-      "loss": 0.4108,
+      "grad_norm": 1.12320077419281,
+      "learning_rate": 0.0004976267678187994,
+      "loss": 0.4391,
       "step": 192
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8052837573385518,
-      "eval_f1": 0.8202348690153567,
-      "eval_loss": 0.39227691292762756,
-      "eval_precision": 0.761744966442953,
-      "eval_recall": 0.8884540117416829,
-      "eval_runtime": 30.7242,
-      "eval_samples_per_second": 33.264,
-      "eval_steps_per_second": 1.042,
+      "eval_accuracy": 0.8082191780821918,
+      "eval_f1": 0.8227848101265822,
+      "eval_loss": 0.41584911942481995,
+      "eval_precision": 0.7647058823529411,
+      "eval_recall": 0.8904109589041096,
+      "eval_runtime": 28.5215,
+      "eval_samples_per_second": 35.833,
+      "eval_steps_per_second": 1.122,
       "step": 192
     },
     {
       "epoch": 3.0,
-      "grad_norm": 2.60581374168396,
-      "learning_rate": 0.00011487016523335267,
-      "loss": 0.3914,
+      "grad_norm": 2.946756362915039,
+      "learning_rate": 0.00033175117854586627,
+      "loss": 0.41,
       "step": 288
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8199608610567515,
-      "eval_f1": 0.8318098720292504,
-      "eval_loss": 0.38357415795326233,
-      "eval_precision": 0.7804459691252144,
-      "eval_recall": 0.8904109589041096,
-      "eval_runtime": 32.3837,
-      "eval_samples_per_second": 31.559,
-      "eval_steps_per_second": 0.988,
+      "eval_accuracy": 0.8140900195694716,
+      "eval_f1": 0.8260073260073261,
+      "eval_loss": 0.43016910552978516,
+      "eval_precision": 0.7762478485370051,
+      "eval_recall": 0.8825831702544031,
+      "eval_runtime": 28.1244,
+      "eval_samples_per_second": 36.339,
+      "eval_steps_per_second": 1.138,
       "step": 288
     },
     {
       "epoch": 4.0,
-      "grad_norm": 3.155740737915039,
-      "learning_rate": 5.743508261667633e-05,
-      "loss": 0.3802,
+      "grad_norm": 3.868119478225708,
+      "learning_rate": 0.00016587558927293314,
+      "loss": 0.4041,
       "step": 384
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8160469667318982,
-      "eval_f1": 0.8160469667318982,
-      "eval_loss": 0.37758293747901917,
-      "eval_precision": 0.8160469667318982,
-      "eval_recall": 0.8160469667318982,
-      "eval_runtime": 30.8515,
-      "eval_samples_per_second": 33.126,
-      "eval_steps_per_second": 1.037,
+      "eval_accuracy": 0.8033268101761253,
+      "eval_f1": 0.802747791952895,
+      "eval_loss": 0.4142443835735321,
+      "eval_precision": 0.8051181102362205,
+      "eval_recall": 0.8003913894324853,
+      "eval_runtime": 29.1758,
+      "eval_samples_per_second": 35.029,
+      "eval_steps_per_second": 1.097,
       "step": 384
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.9643956422805786,
+      "grad_norm": 1.582091212272644,
       "learning_rate": 0.0,
-      "loss": 0.3716,
+      "loss": 0.3858,
       "step": 480
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8277886497064579,
-      "eval_f1": 0.8367346938775511,
-      "eval_loss": 0.37470024824142456,
-      "eval_precision": 0.7954144620811288,
-      "eval_recall": 0.8825831702544031,
-      "eval_runtime": 31.628,
-      "eval_samples_per_second": 32.313,
-      "eval_steps_per_second": 1.012,
+      "eval_accuracy": 0.8238747553816047,
+      "eval_f1": 0.83271375464684,
+      "eval_loss": 0.4042578339576721,
+      "eval_precision": 0.7929203539823009,
+      "eval_recall": 0.8767123287671232,
+      "eval_runtime": 28.2962,
+      "eval_samples_per_second": 36.118,
+      "eval_steps_per_second": 1.131,
       "step": 480
     }
   ],
@@ -113,9 +113,9 @@
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.6818846524061118,
-    "learning_rate": 0.00028717541308338166,
+    "alpha": 0.7818460169984094,
+    "learning_rate": 0.0008293779463646656,
     "num_train_epochs": 5,
-    "temperature": 29
+    "temperature": 3
   }
 }

run-1/checkpoint-480/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:793b34e056c09ac6895e4854810a5a38ad5f7e4ea7bfcdaf2571ad59272936ff
+oid sha256:1ab2cfe4c2ab241510bc23e9ec5727876e620ab477a4e5ad72a937b25e09df23
 size 4920

run-2/checkpoint-96/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:57ef7ac15d70c06615ec62775bbd9d47a7a307f789675ebd91d014018802f0ac
+oid sha256:f72e368a0735e0614c97feb6fba15c5d298f68104e7128cf582de5807e255027
 size 17549312

run-2/checkpoint-96/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8ea9c584a6706cb5bee635fc6f64c59550cbdb80935f1c815de62851495dd156
+oid sha256:5b5aece18803da9b55ab957b14fbd401ac7b58d03d968f7138245dd9246cea4e
 size 35122746

run-2/checkpoint-96/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d5feae9a058dc367ba00ad758a34b0a64ac558c1f8bd5163d9bb2ce9c13aac22
+oid sha256:773b7d3c411900ce077bec534b9aeadeaba63fc284f46150d41278a6fa17bea6
 size 1064

run-2/checkpoint-96/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7984344422700587,
+  "best_metric": 0.5,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-96",
   "epoch": 1.0,
   "eval_steps": 500,
@@ -10,36 +10,36 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 3.4323604106903076,
-      "learning_rate": 0.0004667999543114754,
-      "loss": 0.533,
+      "grad_norm": 3.4041388034820557,
+      "learning_rate": 8.656503841433308e-07,
+      "loss": 0.2746,
       "step": 96
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7984344422700587,
-      "eval_f1": 0.8049242424242424,
-      "eval_loss": 0.4442267119884491,
-      "eval_precision": 0.7798165137614679,
-      "eval_recall": 0.8317025440313112,
-      "eval_runtime": 28.6457,
-      "eval_samples_per_second": 35.677,
-      "eval_steps_per_second": 1.117,
+      "eval_accuracy": 0.5,
+      "eval_f1": 0.6666666666666666,
+      "eval_loss": 0.25451573729515076,
+      "eval_precision": 0.5,
+      "eval_recall": 1.0,
+      "eval_runtime": 28.1178,
+      "eval_samples_per_second": 36.347,
+      "eval_steps_per_second": 1.138,
       "step": 96
     }
   ],
   "logging_steps": 500,
-  "max_steps": 192,
+  "max_steps": 480,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
+  "num_train_epochs": 5,
   "save_steps": 500,
   "total_flos": 235695197280.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.88253398201078,
-    "learning_rate": 0.0009335999086229508,
-    "num_train_epochs": 2,
-    "temperature": 19
+    "alpha": 0.04277939142155329,
+    "learning_rate": 1.0820629801791635e-06,
+    "num_train_epochs": 5,
+    "temperature": 6
   }
 }

run-2/checkpoint-96/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2659042d49fa36ce1404b1f7057dd4643dc22f653d1fc44d521700862149e4b7
+oid sha256:bd5c3fbfe040730ec9bc64036d615ae00f3fe41347ad2df02c2ead7e8faec2da
 size 4920

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ab2cfe4c2ab241510bc23e9ec5727876e620ab477a4e5ad72a937b25e09df23
+oid sha256:bd5c3fbfe040730ec9bc64036d615ae00f3fe41347ad2df02c2ead7e8faec2da
 size 4920