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: 0.3273560876209883
+      value: 0.318246541903987
 ---
 
 <!-- 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. -->
 
 # hBERTv2_new_pretrain_mnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.1000
-- Accuracy: 0.3274
+- Loss: 1.0964
+- Accuracy: 0.3182
 
 ## Model description
 

all_results.json

@@ -1,21 +1,21 @@
 {
-    "epoch": 8.0,
-    "epoch_mm": 8.0,
+    "epoch": 7.0,
+    "epoch_mm": 7.0,
     "eval_accuracy": 0.31818644931227713,
     "eval_accuracy_mm": 0.318246541903987,
-    "eval_loss": 1.0991482734680176,
-    "eval_loss_mm": 1.099164366722107,
-    "eval_runtime": 16.9277,
-    "eval_runtime_mm": 16.9647,
+    "eval_loss": 1.0963791608810425,
+    "eval_loss_mm": 1.0964126586914062,
+    "eval_runtime": 16.6567,
+    "eval_runtime_mm": 16.6682,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 579.818,
-    "eval_samples_per_second_mm": 579.558,
-    "eval_steps_per_second": 4.549,
-    "eval_steps_per_second_mm": 4.539,
-    "train_loss": 11.881152813251202,
-    "train_runtime": 15762.7066,
+    "eval_samples_per_second": 589.253,
+    "eval_samples_per_second_mm": 589.867,
+    "eval_steps_per_second": 4.623,
+    "eval_steps_per_second_mm": 4.62,
+    "train_loss": 1.099463806891268,
+    "train_runtime": 13736.6021,
     "train_samples": 392702,
-    "train_samples_per_second": 1245.668,
-    "train_steps_per_second": 9.732
+    "train_samples_per_second": 1429.4,
+    "train_steps_per_second": 11.167
 }

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 8.0,
-    "epoch_mm": 8.0,
+    "epoch": 7.0,
+    "epoch_mm": 7.0,
     "eval_accuracy": 0.31818644931227713,
     "eval_accuracy_mm": 0.318246541903987,
-    "eval_loss": 1.0991482734680176,
-    "eval_loss_mm": 1.099164366722107,
-    "eval_runtime": 16.9277,
-    "eval_runtime_mm": 16.9647,
+    "eval_loss": 1.0963791608810425,
+    "eval_loss_mm": 1.0964126586914062,
+    "eval_runtime": 16.6567,
+    "eval_runtime_mm": 16.6682,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 579.818,
-    "eval_samples_per_second_mm": 579.558,
-    "eval_steps_per_second": 4.549,
-    "eval_steps_per_second_mm": 4.539
+    "eval_samples_per_second": 589.253,
+    "eval_samples_per_second_mm": 589.867,
+    "eval_steps_per_second": 4.623,
+    "eval_steps_per_second_mm": 4.62
 }

logs/events.out.tfevents.1686068169.serv-3317.3377083.26

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f8653e99dcc1a965816a7f0d732fccc649ceeea6c77a2aaa70965e9798d7db5
+size 698

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 8.0,
-    "train_loss": 11.881152813251202,
-    "train_runtime": 15762.7066,
+    "epoch": 7.0,
+    "train_loss": 1.099463806891268,
+    "train_runtime": 13736.6021,
     "train_samples": 392702,
-    "train_samples_per_second": 1245.668,
-    "train_steps_per_second": 9.732
+    "train_samples_per_second": 1429.4,
+    "train_steps_per_second": 11.167
 }

trainer_state.json

@@ -1,145 +1,130 @@
 {
-  "best_metric": 1.0991482734680176,
-  "best_model_checkpoint": "hBERTv2_new_pretrain_mnli/checkpoint-9204",
-  "epoch": 8.0,
-  "global_step": 24544,
+  "best_metric": 1.0963791608810425,
+  "best_model_checkpoint": "hBERTv2_new_pretrain_mnli/checkpoint-6136",
+  "epoch": 7.0,
+  "global_step": 21476,
   "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": 70.7498,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.1031,
       "step": 3068
     },
     {
       "epoch": 1.0,
       "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 30.0480899810791,
-      "eval_runtime": 16.8669,
-      "eval_samples_per_second": 581.909,
-      "eval_steps_per_second": 4.565,
+      "eval_loss": 1.1007808446884155,
+      "eval_runtime": 16.7433,
+      "eval_samples_per_second": 586.203,
+      "eval_steps_per_second": 4.599,
       "step": 3068
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00048,
-      "loss": 5.5287,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.0989,
       "step": 6136
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.1252200603485107,
-      "eval_runtime": 17.0029,
-      "eval_samples_per_second": 577.254,
-      "eval_steps_per_second": 4.529,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0963791608810425,
+      "eval_runtime": 16.7646,
+      "eval_samples_per_second": 585.459,
+      "eval_steps_per_second": 4.593,
       "step": 6136
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00047,
-      "loss": 1.354,
+      "learning_rate": 3.76e-05,
+      "loss": 1.0987,
       "step": 9204
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.0991482734680176,
-      "eval_runtime": 16.9792,
-      "eval_samples_per_second": 578.06,
-      "eval_steps_per_second": 4.535,
+      "eval_accuracy": 0.32654100866021396,
+      "eval_loss": 1.0979726314544678,
+      "eval_runtime": 16.7522,
+      "eval_samples_per_second": 585.894,
+      "eval_steps_per_second": 4.596,
       "step": 9204
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00046,
-      "loss": 3.0766,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 1.0987,
       "step": 12272
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.2373435497283936,
-      "eval_runtime": 16.9102,
-      "eval_samples_per_second": 580.418,
-      "eval_steps_per_second": 4.553,
+      "eval_loss": 1.0994985103607178,
+      "eval_runtime": 16.749,
+      "eval_samples_per_second": 586.006,
+      "eval_steps_per_second": 4.597,
       "step": 12272
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.00045000000000000004,
-      "loss": 4.2409,
+      "learning_rate": 3.6e-05,
+      "loss": 1.0986,
       "step": 15340
     },
     {
       "epoch": 5.0,
       "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.191184163093567,
-      "eval_runtime": 16.9866,
-      "eval_samples_per_second": 577.809,
-      "eval_steps_per_second": 4.533,
+      "eval_loss": 1.0985783338546753,
+      "eval_runtime": 16.7396,
+      "eval_samples_per_second": 586.335,
+      "eval_steps_per_second": 4.6,
       "step": 15340
     },
     {
       "epoch": 6.0,
-      "learning_rate": 0.00044,
-      "loss": 5.4957,
+      "learning_rate": 3.52e-05,
+      "loss": 1.0991,
       "step": 18408
     },
     {
       "epoch": 6.0,
       "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.1716059446334839,
-      "eval_runtime": 16.9674,
-      "eval_samples_per_second": 578.463,
-      "eval_steps_per_second": 4.538,
+      "eval_loss": 1.1015625,
+      "eval_runtime": 16.7478,
+      "eval_samples_per_second": 586.047,
+      "eval_steps_per_second": 4.598,
       "step": 18408
     },
     {
       "epoch": 7.0,
-      "learning_rate": 0.00043,
-      "loss": 3.0327,
+      "learning_rate": 3.44e-05,
+      "loss": 1.0991,
       "step": 21476
     },
     {
       "epoch": 7.0,
       "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.315549373626709,
-      "eval_runtime": 16.8941,
-      "eval_samples_per_second": 580.971,
-      "eval_steps_per_second": 4.558,
+      "eval_loss": 1.099992036819458,
+      "eval_runtime": 16.7418,
+      "eval_samples_per_second": 586.259,
+      "eval_steps_per_second": 4.599,
       "step": 21476
     },
     {
-      "epoch": 8.0,
-      "learning_rate": 0.00042,
-      "loss": 1.571,
-      "step": 24544
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.3787474632263184,
-      "eval_runtime": 16.9205,
-      "eval_samples_per_second": 580.066,
-      "eval_steps_per_second": 4.551,
-      "step": 24544
-    },
-    {
-      "epoch": 8.0,
-      "step": 24544,
-      "total_flos": 4.597451878333153e+17,
-      "train_loss": 11.881152813251202,
-      "train_runtime": 15762.7066,
-      "train_samples_per_second": 1245.668,
-      "train_steps_per_second": 9.732
+      "epoch": 7.0,
+      "step": 21476,
+      "total_flos": 4.022770393541509e+17,
+      "train_loss": 1.099463806891268,
+      "train_runtime": 13736.6021,
+      "train_samples_per_second": 1429.4,
+      "train_steps_per_second": 11.167
     }
   ],
   "max_steps": 153400,
   "num_train_epochs": 50,
-  "total_flos": 4.597451878333153e+17,
+  "total_flos": 4.022770393541509e+17,
   "trial_name": null,
   "trial_params": null
 }