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 WNLI
       type: glue
       config: wnli
       split: validation
@@ -20,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.4647887323943662
+      value: 0.5633802816901409
 ---
 
 <!-- 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_wnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new) on the GLUE WNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6922
-- Accuracy: 0.4648
+- Loss: 0.6852
+- Accuracy: 0.5634
 
 ## Model description
 

all_results.json

@@ -1,14 +1,14 @@
 {
-    "epoch": 6.0,
-    "eval_accuracy": 0.43661971830985913,
-    "eval_loss": 30.190143585205078,
-    "eval_runtime": 0.1401,
+    "epoch": 12.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.6852443218231201,
+    "eval_runtime": 0.1404,
     "eval_samples": 71,
-    "eval_samples_per_second": 506.621,
-    "eval_steps_per_second": 7.136,
-    "train_loss": 29.312188212076823,
-    "train_runtime": 42.2828,
+    "eval_samples_per_second": 505.684,
+    "eval_steps_per_second": 7.122,
+    "train_loss": 0.7179187575976054,
+    "train_runtime": 110.2125,
     "train_samples": 635,
-    "train_samples_per_second": 750.897,
-    "train_steps_per_second": 5.913
+    "train_samples_per_second": 288.08,
+    "train_steps_per_second": 2.268
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 6.0,
-    "eval_accuracy": 0.43661971830985913,
-    "eval_loss": 30.190143585205078,
-    "eval_runtime": 0.1401,
+    "epoch": 12.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.6852443218231201,
+    "eval_runtime": 0.1404,
     "eval_samples": 71,
-    "eval_samples_per_second": 506.621,
-    "eval_steps_per_second": 7.136
+    "eval_samples_per_second": 505.684,
+    "eval_steps_per_second": 7.122
 }

logs/events.out.tfevents.1686054137.serv-3317.3374904.23

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1f9c49297096b898fe426a8feec0e0759021445b14e6fabd3247a20ecfd8985
+size 357

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 6.0,
-    "train_loss": 29.312188212076823,
-    "train_runtime": 42.2828,
+    "epoch": 12.0,
+    "train_loss": 0.7179187575976054,
+    "train_runtime": 110.2125,
     "train_samples": 635,
-    "train_samples_per_second": 750.897,
-    "train_steps_per_second": 5.913
+    "train_samples_per_second": 288.08,
+    "train_steps_per_second": 2.268
 }

trainer_state.json

@@ -1,115 +1,205 @@
 {
-  "best_metric": 30.190143585205078,
-  "best_model_checkpoint": "hBERTv1_new_pretrain_wnli/checkpoint-5",
-  "epoch": 6.0,
-  "global_step": 30,
+  "best_metric": 0.6852443218231201,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_wnli/checkpoint-35",
+  "epoch": 12.0,
+  "global_step": 60,
   "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": 29.5316,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.8538,
       "step": 5
     },
     {
       "epoch": 1.0,
       "eval_accuracy": 0.43661971830985913,
-      "eval_loss": 30.190143585205078,
-      "eval_runtime": 0.1384,
-      "eval_samples_per_second": 513.028,
-      "eval_steps_per_second": 7.226,
+      "eval_loss": 0.697458028793335,
+      "eval_runtime": 0.1385,
+      "eval_samples_per_second": 512.476,
+      "eval_steps_per_second": 7.218,
       "step": 5
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00048,
-      "loss": 29.2255,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.7194,
       "step": 10
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.43661971830985913,
-      "eval_loss": 30.190143585205078,
-      "eval_runtime": 0.1388,
-      "eval_samples_per_second": 511.62,
-      "eval_steps_per_second": 7.206,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.6921765208244324,
+      "eval_runtime": 0.1386,
+      "eval_samples_per_second": 512.215,
+      "eval_steps_per_second": 7.214,
       "step": 10
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00047,
-      "loss": 28.1924,
+      "learning_rate": 3.76e-05,
+      "loss": 0.7223,
       "step": 15
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.43661971830985913,
-      "eval_loss": 30.190143585205078,
-      "eval_runtime": 0.1374,
-      "eval_samples_per_second": 516.831,
-      "eval_steps_per_second": 7.279,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.6893155574798584,
+      "eval_runtime": 0.1342,
+      "eval_samples_per_second": 529.105,
+      "eval_steps_per_second": 7.452,
       "step": 15
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00046,
-      "loss": 30.1423,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.713,
       "step": 20
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.43661971830985913,
-      "eval_loss": 30.190143585205078,
-      "eval_runtime": 0.1372,
-      "eval_samples_per_second": 517.375,
-      "eval_steps_per_second": 7.287,
+      "eval_loss": 0.720455527305603,
+      "eval_runtime": 0.136,
+      "eval_samples_per_second": 521.953,
+      "eval_steps_per_second": 7.351,
       "step": 20
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.00045000000000000004,
-      "loss": 29.1643,
+      "learning_rate": 3.6e-05,
+      "loss": 0.7081,
       "step": 25
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.43661971830985913,
-      "eval_loss": 30.190143585205078,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.686454713344574,
       "eval_runtime": 0.1356,
-      "eval_samples_per_second": 523.545,
-      "eval_steps_per_second": 7.374,
+      "eval_samples_per_second": 523.488,
+      "eval_steps_per_second": 7.373,
       "step": 25
     },
     {
       "epoch": 6.0,
-      "learning_rate": 0.00044,
-      "loss": 29.6169,
+      "learning_rate": 3.52e-05,
+      "loss": 0.7028,
       "step": 30
     },
     {
       "epoch": 6.0,
       "eval_accuracy": 0.43661971830985913,
-      "eval_loss": 30.190143585205078,
-      "eval_runtime": 0.1362,
-      "eval_samples_per_second": 521.15,
-      "eval_steps_per_second": 7.34,
+      "eval_loss": 0.7047755718231201,
+      "eval_runtime": 0.1352,
+      "eval_samples_per_second": 525.033,
+      "eval_steps_per_second": 7.395,
       "step": 30
     },
     {
-      "epoch": 6.0,
-      "step": 30,
-      "total_flos": 564248728043520.0,
-      "train_loss": 29.312188212076823,
-      "train_runtime": 42.2828,
-      "train_samples_per_second": 750.897,
-      "train_steps_per_second": 5.913
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.697,
+      "step": 35
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.6852443218231201,
+      "eval_runtime": 0.1355,
+      "eval_samples_per_second": 523.974,
+      "eval_steps_per_second": 7.38,
+      "step": 35
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.7002,
+      "step": 40
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.6967429518699646,
+      "eval_runtime": 0.139,
+      "eval_samples_per_second": 510.953,
+      "eval_steps_per_second": 7.197,
+      "step": 40
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.28e-05,
+      "loss": 0.7017,
+      "step": 45
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.7156140208244324,
+      "eval_runtime": 0.1354,
+      "eval_samples_per_second": 524.369,
+      "eval_steps_per_second": 7.385,
+      "step": 45
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.702,
+      "step": 50
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.688545286655426,
+      "eval_runtime": 0.1368,
+      "eval_samples_per_second": 519.171,
+      "eval_steps_per_second": 7.312,
+      "step": 50
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.1200000000000006e-05,
+      "loss": 0.6945,
+      "step": 55
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.49295774647887325,
+      "eval_loss": 0.6926718354225159,
+      "eval_runtime": 0.1366,
+      "eval_samples_per_second": 519.835,
+      "eval_steps_per_second": 7.322,
+      "step": 55
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 3.0400000000000004e-05,
+      "loss": 0.7002,
+      "step": 60
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.4647887323943662,
+      "eval_loss": 0.6922314763069153,
+      "eval_runtime": 0.1368,
+      "eval_samples_per_second": 519.164,
+      "eval_steps_per_second": 7.312,
+      "step": 60
+    },
+    {
+      "epoch": 12.0,
+      "step": 60,
+      "total_flos": 1128497456087040.0,
+      "train_loss": 0.7179187575976054,
+      "train_runtime": 110.2125,
+      "train_samples_per_second": 288.08,
+      "train_steps_per_second": 2.268
     }
   ],
   "max_steps": 250,
   "num_train_epochs": 50,
-  "total_flos": 564248728043520.0,
+  "total_flos": 1128497456087040.0,
   "trial_name": null,
   "trial_params": null
 }