Training in progress, epoch 4

logs/events.out.tfevents.1709712823.a2333cf40ebd.21010.53

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d953dd5693ad21811d1fd196355dd096862dc8c5967e02a2ec9339794caa6d0
+size 7364

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9aad9c62b38123d39495e073be4846edc80536196a5257982b3af68214c186d6
+oid sha256:7a5cced5d1408a3231d9adec89e92636cb688ecf5bcd1f99c9679e57f7106218
 size 17549312

run-0/checkpoint-192/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a7c5e94d351fdbc0935e7dbe3174244feb7ddadc5f63dc2f16808c51b2af2a57
+oid sha256:e8866e4d6c1015be008c4b6e31b26cd77f7e7ebe9f1ff486a3715bbdf96b3fb5
 size 17549312

run-0/checkpoint-192/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ac78cbe4519b949fd207235051b160ea7f7956b055fb088955c4a1ac6b0df268
+oid sha256:6d80f7532747c0db7352fe4cff783b23f284706c5544e11e300ee7bb139bafc9
 size 35122746

run-0/checkpoint-192/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:13915667501739529d666ffe624aa321331ccbffecf0ae7d04a789ecdbde4536
+oid sha256:911745f0e5474343f42a33d0a58e92d24ad5d65d3db76f347f1d80c1338d3ba2
 size 1064

run-0/checkpoint-192/tokenizer.json

@@ -2,7 +2,7 @@
   "version": "1.0",
   "truncation": {
     "direction": "Right",
-    "max_length": 31,
+    "max_length": 29,
     "strategy": "LongestFirst",
     "stride": 0
   },

run-0/checkpoint-192/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7896281800391389,
+  "best_metric": 0.5136986301369864,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-192",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,40 +10,40 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 1.8004393577575684,
-      "learning_rate": 0.0007697157586205755,
-      "loss": 0.4935,
+      "grad_norm": 0.8228768110275269,
+      "learning_rate": 1.922542063523116e-05,
+      "loss": 0.7149,
       "step": 96
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7436399217221135,
-      "eval_f1": 0.7689594356261024,
-      "eval_loss": 0.448881596326828,
-      "eval_precision": 0.6998394863563403,
-      "eval_recall": 0.8532289628180039,
-      "eval_runtime": 28.6476,
-      "eval_samples_per_second": 35.675,
-      "eval_steps_per_second": 1.117,
+      "eval_accuracy": 0.5,
+      "eval_f1": 0.007766990291262136,
+      "eval_loss": 0.7131175398826599,
+      "eval_precision": 0.5,
+      "eval_recall": 0.003913894324853229,
+      "eval_runtime": 26.5316,
+      "eval_samples_per_second": 38.52,
+      "eval_steps_per_second": 1.206,
       "step": 96
     },
     {
       "epoch": 2.0,
-      "grad_norm": 4.9879961013793945,
-      "learning_rate": 0.0006841917854405115,
-      "loss": 0.4571,
+      "grad_norm": 2.1306586265563965,
+      "learning_rate": 1.7089262786872146e-05,
+      "loss": 0.7062,
       "step": 192
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7896281800391389,
-      "eval_f1": 0.8057813911472448,
-      "eval_loss": 0.42557916045188904,
-      "eval_precision": 0.7483221476510067,
-      "eval_recall": 0.87279843444227,
-      "eval_runtime": 28.772,
-      "eval_samples_per_second": 35.521,
-      "eval_steps_per_second": 1.112,
+      "eval_accuracy": 0.5136986301369864,
+      "eval_f1": 0.08133086876155268,
+      "eval_loss": 0.7069234848022461,
+      "eval_precision": 0.7333333333333333,
+      "eval_recall": 0.043052837573385516,
+      "eval_runtime": 26.8847,
+      "eval_samples_per_second": 38.014,
+      "eval_steps_per_second": 1.19,
       "step": 192
     }
   ],
@@ -52,13 +52,13 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
-  "total_flos": 471390394560.0,
+  "total_flos": 440978111040.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.735211359802956,
-    "learning_rate": 0.0008552397318006393,
+    "alpha": 0.8789746577248793,
+    "learning_rate": 2.136157848359018e-05,
     "num_train_epochs": 10,
-    "temperature": 25
+    "temperature": 14
   }
 }

run-0/checkpoint-192/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:23804eaaa0feed30486319d2982557f65dce708f7d788864c811c20b2e8c33b4
+oid sha256:bf1d86d1ef9c5f73f9fb51b0ea325423f470c0f4f628eb656cb6d38e041dfed8
 size 4920

run-0/checkpoint-384/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1c6b74a4943e5a0cd97e5e1c9bf5294e2e984ae36031447242ee7cc0ed3cb215
+oid sha256:7a5cced5d1408a3231d9adec89e92636cb688ecf5bcd1f99c9679e57f7106218
 size 17549312

run-0/checkpoint-384/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e0f93f2ccccb17743dfff87bebe675e197173b445069b6bdab70e7ae8cae17ca
+oid sha256:c0e3db2ea3c5d266e5b6785e41f67a9da15b4493f25e13329b02022502d9e2e7
 size 35122746

run-0/checkpoint-384/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:909efe5e52b158b9db7198ea1df0d8ef2b972c80ecedca1230a879e487e4aac0
+oid sha256:a2efd5220d80201e72fefaf32ad5da346b0aa9bc0a3be9b0c7b351c23430409f
 size 1064

run-0/checkpoint-384/tokenizer.json

@@ -2,7 +2,7 @@
   "version": "1.0",
   "truncation": {
     "direction": "Right",
-    "max_length": 31,
+    "max_length": 29,
     "strategy": "LongestFirst",
     "stride": 0
   },

run-0/checkpoint-384/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7377690802348337,
+  "best_metric": 0.5606653620352251,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-384",
   "epoch": 4.0,
   "eval_steps": 500,
@@ -10,93 +10,93 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 0.4804088771343231,
-      "learning_rate": 2.9818131608660274e-05,
-      "loss": 0.4786,
+      "grad_norm": 0.8228768110275269,
+      "learning_rate": 1.922542063523116e-05,
+      "loss": 0.7149,
       "step": 96
     },
     {
       "epoch": 1.0,
       "eval_accuracy": 0.5,
-      "eval_f1": 0.0,
-      "eval_loss": 0.4664871394634247,
-      "eval_precision": 0.0,
-      "eval_recall": 0.0,
-      "eval_runtime": 28.9797,
-      "eval_samples_per_second": 35.266,
-      "eval_steps_per_second": 1.104,
+      "eval_f1": 0.007766990291262136,
+      "eval_loss": 0.7131175398826599,
+      "eval_precision": 0.5,
+      "eval_recall": 0.003913894324853229,
+      "eval_runtime": 26.5316,
+      "eval_samples_per_second": 38.52,
+      "eval_steps_per_second": 1.206,
       "step": 96
     },
     {
       "epoch": 2.0,
-      "grad_norm": 1.77193021774292,
-      "learning_rate": 1.9878754405773516e-05,
-      "loss": 0.4481,
+      "grad_norm": 2.1306586265563965,
+      "learning_rate": 1.7089262786872146e-05,
+      "loss": 0.7062,
       "step": 192
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5958904109589042,
-      "eval_f1": 0.37707390648567113,
-      "eval_loss": 0.4186549484729767,
-      "eval_precision": 0.8223684210526315,
-      "eval_recall": 0.2446183953033268,
-      "eval_runtime": 28.789,
-      "eval_samples_per_second": 35.5,
-      "eval_steps_per_second": 1.112,
+      "eval_accuracy": 0.5136986301369864,
+      "eval_f1": 0.08133086876155268,
+      "eval_loss": 0.7069234848022461,
+      "eval_precision": 0.7333333333333333,
+      "eval_recall": 0.043052837573385516,
+      "eval_runtime": 26.8847,
+      "eval_samples_per_second": 38.014,
+      "eval_steps_per_second": 1.19,
       "step": 192
     },
     {
       "epoch": 3.0,
-      "grad_norm": 1.5943424701690674,
-      "learning_rate": 9.939377202886758e-06,
-      "loss": 0.4139,
+      "grad_norm": 1.5551536083221436,
+      "learning_rate": 1.4953104938513126e-05,
+      "loss": 0.6972,
       "step": 288
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.6790606653620352,
-      "eval_f1": 0.6019417475728155,
-      "eval_loss": 0.39717963337898254,
-      "eval_precision": 0.792332268370607,
-      "eval_recall": 0.48532289628180036,
-      "eval_runtime": 29.1776,
-      "eval_samples_per_second": 35.027,
-      "eval_steps_per_second": 1.097,
+      "eval_accuracy": 0.512720156555773,
+      "eval_f1": 0.08791208791208792,
+      "eval_loss": 0.6998902559280396,
+      "eval_precision": 0.6857142857142857,
+      "eval_recall": 0.046966731898238745,
+      "eval_runtime": 26.6683,
+      "eval_samples_per_second": 38.323,
+      "eval_steps_per_second": 1.2,
       "step": 288
     },
     {
       "epoch": 4.0,
-      "grad_norm": 1.1352671384811401,
-      "learning_rate": 0.0,
-      "loss": 0.4045,
+      "grad_norm": 2.080883741378784,
+      "learning_rate": 1.2816947090154107e-05,
+      "loss": 0.6854,
       "step": 384
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.7377690802348337,
-      "eval_f1": 0.7184873949579831,
-      "eval_loss": 0.392328679561615,
-      "eval_precision": 0.7755102040816326,
-      "eval_recall": 0.6692759295499021,
-      "eval_runtime": 28.937,
-      "eval_samples_per_second": 35.318,
-      "eval_steps_per_second": 1.106,
+      "eval_accuracy": 0.5606653620352251,
+      "eval_f1": 0.3145038167938931,
+      "eval_loss": 0.6807624697685242,
+      "eval_precision": 0.7152777777777778,
+      "eval_recall": 0.20156555772994128,
+      "eval_runtime": 26.6572,
+      "eval_samples_per_second": 38.339,
+      "eval_steps_per_second": 1.2,
       "step": 384
     }
   ],
   "logging_steps": 500,
-  "max_steps": 384,
+  "max_steps": 960,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 10,
   "save_steps": 500,
-  "total_flos": 942780789120.0,
+  "total_flos": 881956222080.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.6186723691968887,
-    "learning_rate": 3.975750881154703e-05,
-    "num_train_epochs": 4,
-    "temperature": 6
+    "alpha": 0.8789746577248793,
+    "learning_rate": 2.136157848359018e-05,
+    "num_train_epochs": 10,
+    "temperature": 14
   }
 }

run-0/checkpoint-384/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3aa29ea69dd18a01982daa5b87211d303a5144b609fb7e84569d6d408da8e995
+oid sha256:bf1d86d1ef9c5f73f9fb51b0ea325423f470c0f4f628eb656cb6d38e041dfed8
 size 4920

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:019a0c30743454578a20e039f50246add333dec6abb36273bd6bc2339b797b34
+oid sha256:bf1d86d1ef9c5f73f9fb51b0ea325423f470c0f4f628eb656cb6d38e041dfed8
 size 4920