End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 tags:
 - generated_from_trainer
 datasets:
@@ -12,7 +14,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE MNLI
       type: glue
       config: mnli
       split: validation_matched
@@ -20,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.31818644931227713
+      value: 0.36421887713588286
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -28,10 +30,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hBERTv1_new_pretrain_w_init__mnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.1005
-- Accuracy: 0.3182
+- Loss: 1.0846
+- Accuracy: 0.3642
 
 ## Model description
 

all_results.json

@@ -1,21 +1,21 @@
 {
-    "epoch": 7.0,
-    "epoch_mm": 7.0,
-    "eval_accuracy": 0.31818644931227713,
-    "eval_accuracy_mm": 0.318246541903987,
-    "eval_loss": 1.0963075160980225,
-    "eval_loss_mm": 1.0963244438171387,
-    "eval_runtime": 17.5761,
-    "eval_runtime_mm": 17.6104,
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.3515028018339277,
+    "eval_accuracy_mm": 0.36421887713588286,
+    "eval_loss": 1.0895687341690063,
+    "eval_loss_mm": 1.084586262702942,
+    "eval_runtime": 17.5312,
+    "eval_runtime_mm": 17.5748,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 558.429,
-    "eval_samples_per_second_mm": 558.305,
-    "eval_steps_per_second": 4.381,
-    "eval_steps_per_second_mm": 4.372,
-    "train_loss": 1.1130940406587586,
-    "train_runtime": 14060.0313,
+    "eval_samples_per_second": 559.861,
+    "eval_samples_per_second_mm": 559.436,
+    "eval_steps_per_second": 4.392,
+    "eval_steps_per_second_mm": 4.381,
+    "train_loss": 1.09808146689364,
+    "train_runtime": 20906.5595,
     "train_samples": 392702,
-    "train_samples_per_second": 1396.519,
-    "train_steps_per_second": 10.91
+    "train_samples_per_second": 939.184,
+    "train_steps_per_second": 7.337
 }

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 7.0,
-    "epoch_mm": 7.0,
-    "eval_accuracy": 0.31818644931227713,
-    "eval_accuracy_mm": 0.318246541903987,
-    "eval_loss": 1.0963075160980225,
-    "eval_loss_mm": 1.0963244438171387,
-    "eval_runtime": 17.5761,
-    "eval_runtime_mm": 17.6104,
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.3515028018339277,
+    "eval_accuracy_mm": 0.36421887713588286,
+    "eval_loss": 1.0895687341690063,
+    "eval_loss_mm": 1.084586262702942,
+    "eval_runtime": 17.5312,
+    "eval_runtime_mm": 17.5748,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 558.429,
-    "eval_samples_per_second_mm": 558.305,
-    "eval_steps_per_second": 4.381,
-    "eval_steps_per_second_mm": 4.372
+    "eval_samples_per_second": 559.861,
+    "eval_samples_per_second_mm": 559.436,
+    "eval_steps_per_second": 4.392,
+    "eval_steps_per_second_mm": 4.381
 }

logs/events.out.tfevents.1686079018.serv-3317.3377907.26

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba10b41ec3f867daefe2f8883faa0edfc68f416db4cf08abc7bb3fb0ccaf7642
+size 698

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 7.0,
-    "train_loss": 1.1130940406587586,
-    "train_runtime": 14060.0313,
+    "epoch": 10.0,
+    "train_loss": 1.09808146689364,
+    "train_runtime": 20906.5595,
     "train_samples": 392702,
-    "train_samples_per_second": 1396.519,
-    "train_steps_per_second": 10.91
+    "train_samples_per_second": 939.184,
+    "train_steps_per_second": 7.337
 }

trainer_state.json

@@ -1,130 +1,175 @@
 {
-  "best_metric": 1.0963075160980225,
-  "best_model_checkpoint": "hBERTv1_new_pretrain_w_init__mnli/checkpoint-6136",
-  "epoch": 7.0,
-  "global_step": 21476,
+  "best_metric": 1.0895687341690063,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_w_init__mnli/checkpoint-15340",
+  "epoch": 10.0,
+  "global_step": 30680,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "learning_rate": 0.00049,
-      "loss": 1.1346,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.0911,
       "step": 3068
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0987932682037354,
-      "eval_runtime": 17.7369,
-      "eval_samples_per_second": 553.367,
-      "eval_steps_per_second": 4.341,
+      "eval_accuracy": 0.33397860417727965,
+      "eval_loss": 1.094429612159729,
+      "eval_runtime": 17.6447,
+      "eval_samples_per_second": 556.258,
+      "eval_steps_per_second": 4.364,
       "step": 3068
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00048,
-      "loss": 1.1013,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.0984,
       "step": 6136
     },
     {
       "epoch": 2.0,
       "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.0963075160980225,
-      "eval_runtime": 17.7358,
-      "eval_samples_per_second": 553.402,
-      "eval_steps_per_second": 4.342,
+      "eval_loss": 1.096695065498352,
+      "eval_runtime": 17.6375,
+      "eval_samples_per_second": 556.485,
+      "eval_steps_per_second": 4.366,
       "step": 6136
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00047,
-      "loss": 1.1611,
+      "learning_rate": 3.76e-05,
+      "loss": 1.0988,
       "step": 9204
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0990049839019775,
-      "eval_runtime": 17.7014,
-      "eval_samples_per_second": 554.477,
-      "eval_steps_per_second": 4.35,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.0961673259735107,
+      "eval_runtime": 17.6642,
+      "eval_samples_per_second": 555.643,
+      "eval_steps_per_second": 4.359,
       "step": 9204
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00046,
-      "loss": 1.0988,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 1.0989,
       "step": 12272
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0987932682037354,
-      "eval_runtime": 17.7275,
-      "eval_samples_per_second": 553.658,
-      "eval_steps_per_second": 4.344,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.097014307975769,
+      "eval_runtime": 17.6481,
+      "eval_samples_per_second": 556.152,
+      "eval_steps_per_second": 4.363,
       "step": 12272
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.00045000000000000004,
-      "loss": 1.0987,
+      "learning_rate": 3.6e-05,
+      "loss": 1.0986,
       "step": 15340
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0987932682037354,
-      "eval_runtime": 17.7181,
-      "eval_samples_per_second": 553.953,
-      "eval_steps_per_second": 4.346,
+      "eval_accuracy": 0.3515028018339277,
+      "eval_loss": 1.0895687341690063,
+      "eval_runtime": 17.6444,
+      "eval_samples_per_second": 556.267,
+      "eval_steps_per_second": 4.364,
       "step": 15340
     },
     {
       "epoch": 6.0,
-      "learning_rate": 0.00044,
-      "loss": 1.0985,
+      "learning_rate": 3.52e-05,
+      "loss": 1.0984,
       "step": 18408
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0990049839019775,
-      "eval_runtime": 17.7077,
-      "eval_samples_per_second": 554.278,
-      "eval_steps_per_second": 4.348,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.098585605621338,
+      "eval_runtime": 17.6522,
+      "eval_samples_per_second": 556.022,
+      "eval_steps_per_second": 4.362,
       "step": 18408
     },
     {
       "epoch": 7.0,
-      "learning_rate": 0.00043,
-      "loss": 1.0987,
+      "learning_rate": 3.44e-05,
+      "loss": 1.0992,
       "step": 21476
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0990049839019775,
-      "eval_runtime": 17.6862,
-      "eval_samples_per_second": 554.953,
-      "eval_steps_per_second": 4.354,
+      "eval_loss": 1.0993512868881226,
+      "eval_runtime": 17.6446,
+      "eval_samples_per_second": 556.262,
+      "eval_steps_per_second": 4.364,
       "step": 21476
     },
     {
-      "epoch": 7.0,
-      "step": 21476,
-      "total_flos": 4.071085382597345e+17,
-      "train_loss": 1.1130940406587586,
-      "train_runtime": 14060.0313,
-      "train_samples_per_second": 1396.519,
-      "train_steps_per_second": 10.91
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.0993,
+      "step": 24544
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.099851131439209,
+      "eval_runtime": 17.6553,
+      "eval_samples_per_second": 555.922,
+      "eval_steps_per_second": 4.361,
+      "step": 24544
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.28e-05,
+      "loss": 1.099,
+      "step": 27612
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0964539051055908,
+      "eval_runtime": 17.6517,
+      "eval_samples_per_second": 556.037,
+      "eval_steps_per_second": 4.362,
+      "step": 27612
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.0991,
+      "step": 30680
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.1004576683044434,
+      "eval_runtime": 17.6478,
+      "eval_samples_per_second": 556.161,
+      "eval_steps_per_second": 4.363,
+      "step": 30680
+    },
+    {
+      "epoch": 10.0,
+      "step": 30680,
+      "total_flos": 5.81583626085335e+17,
+      "train_loss": 1.09808146689364,
+      "train_runtime": 20906.5595,
+      "train_samples_per_second": 939.184,
+      "train_steps_per_second": 7.337
     }
   ],
   "max_steps": 153400,
   "num_train_epochs": 50,
-  "total_flos": 4.071085382597345e+17,
+  "total_flos": 5.81583626085335e+17,
   "trial_name": null,
   "trial_params": null
 }