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: 0.3522172497965826
 ---
 
 <!-- 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. -->
 
 # add_BERT_no_pretrain_mnli
 
-This model is a fine-tuned version of [](https://huggingface.co/) on the glue dataset.
+This model is a fine-tuned version of [](https://huggingface.co/) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0986
-- Accuracy: 0.3274
+- Loss: 1.0984
+- Accuracy: 0.3522
 
 ## Model description
 

all_results.json

@@ -1,21 +1,21 @@
 {
-    "epoch": 9.0,
-    "epoch_mm": 9.0,
+    "epoch": 8.0,
+    "epoch_mm": 8.0,
     "eval_accuracy": 0.3544574630667346,
     "eval_accuracy_mm": 0.3522172497965826,
-    "eval_loss": 1.0981897115707397,
-    "eval_loss_mm": 1.0982112884521484,
-    "eval_runtime": 15.2403,
-    "eval_runtime_mm": 15.2786,
+    "eval_loss": 1.0984020233154297,
+    "eval_loss_mm": 1.0984286069869995,
+    "eval_runtime": 18.1962,
+    "eval_runtime_mm": 18.2805,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 644.018,
-    "eval_samples_per_second_mm": 643.512,
-    "eval_steps_per_second": 5.052,
-    "eval_steps_per_second_mm": 5.04,
-    "train_loss": 1.0999290425415127,
-    "train_runtime": 15619.8279,
+    "eval_samples_per_second": 539.398,
+    "eval_samples_per_second_mm": 537.841,
+    "eval_steps_per_second": 4.232,
+    "eval_steps_per_second_mm": 4.212,
+    "train_loss": 1.0992557968278907,
+    "train_runtime": 17750.054,
     "train_samples": 392702,
-    "train_samples_per_second": 1257.063,
-    "train_steps_per_second": 9.821
+    "train_samples_per_second": 1106.199,
+    "train_steps_per_second": 8.642
 }

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 9.0,
-    "epoch_mm": 9.0,
+    "epoch": 8.0,
+    "epoch_mm": 8.0,
     "eval_accuracy": 0.3544574630667346,
     "eval_accuracy_mm": 0.3522172497965826,
-    "eval_loss": 1.0981897115707397,
-    "eval_loss_mm": 1.0982112884521484,
-    "eval_runtime": 15.2403,
-    "eval_runtime_mm": 15.2786,
+    "eval_loss": 1.0984020233154297,
+    "eval_loss_mm": 1.0984286069869995,
+    "eval_runtime": 18.1962,
+    "eval_runtime_mm": 18.2805,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 644.018,
-    "eval_samples_per_second_mm": 643.512,
-    "eval_steps_per_second": 5.052,
-    "eval_steps_per_second_mm": 5.04
+    "eval_samples_per_second": 539.398,
+    "eval_samples_per_second_mm": 537.841,
+    "eval_steps_per_second": 4.232,
+    "eval_steps_per_second_mm": 4.212
 }

logs/events.out.tfevents.1686796314.serv-9219.1715196.17

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6648e4cd32a27aaf8d4d65832f6577f23240a9b6acd75a51c324bd4bbb29491b
+size 698

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.0,
-    "train_loss": 1.0999290425415127,
-    "train_runtime": 15619.8279,
+    "epoch": 8.0,
+    "train_loss": 1.0992557968278907,
+    "train_runtime": 17750.054,
     "train_samples": 392702,
-    "train_samples_per_second": 1257.063,
-    "train_steps_per_second": 9.821
+    "train_samples_per_second": 1106.199,
+    "train_steps_per_second": 8.642
 }

trainer_state.json

@@ -1,160 +1,145 @@
 {
-  "best_metric": 1.0981897115707397,
-  "best_model_checkpoint": "add_BERT_no_pretrain_mnli/checkpoint-12272",
-  "epoch": 9.0,
-  "global_step": 27612,
+  "best_metric": 1.0984020233154297,
+  "best_model_checkpoint": "add_BERT_no_pretrain_mnli/checkpoint-9204",
+  "epoch": 8.0,
+  "global_step": 24544,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "learning_rate": 0.000490006518904824,
-      "loss": 1.1062,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.1024,
       "step": 3068
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0982247591018677,
-      "eval_runtime": 15.2821,
-      "eval_samples_per_second": 642.254,
-      "eval_steps_per_second": 5.039,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0987340211868286,
+      "eval_runtime": 18.3205,
+      "eval_samples_per_second": 535.74,
+      "eval_steps_per_second": 4.203,
       "step": 3068
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.000480006518904824,
-      "loss": 1.0987,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.0997,
       "step": 6136
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.0985891819000244,
-      "eval_runtime": 15.3034,
-      "eval_samples_per_second": 641.359,
-      "eval_steps_per_second": 5.032,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.0985723733901978,
+      "eval_runtime": 18.3326,
+      "eval_samples_per_second": 535.385,
+      "eval_steps_per_second": 4.2,
       "step": 6136
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.000470009778357236,
-      "loss": 1.0987,
+      "learning_rate": 3.76e-05,
+      "loss": 1.0988,
       "step": 9204
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0984275341033936,
-      "eval_runtime": 15.2993,
-      "eval_samples_per_second": 641.531,
-      "eval_steps_per_second": 5.033,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.0984020233154297,
+      "eval_runtime": 18.3176,
+      "eval_samples_per_second": 535.823,
+      "eval_steps_per_second": 4.204,
       "step": 9204
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.000460009778357236,
-      "loss": 1.0986,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 1.0987,
       "step": 12272
     },
     {
       "epoch": 4.0,
       "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0981897115707397,
-      "eval_runtime": 15.3289,
-      "eval_samples_per_second": 640.295,
-      "eval_steps_per_second": 5.023,
+      "eval_loss": 1.0986018180847168,
+      "eval_runtime": 18.4288,
+      "eval_samples_per_second": 532.589,
+      "eval_steps_per_second": 4.178,
       "step": 12272
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.00045001629726206,
+      "learning_rate": 3.6e-05,
       "loss": 1.0986,
       "step": 15340
     },
     {
       "epoch": 5.0,
       "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0984646081924438,
-      "eval_runtime": 15.31,
-      "eval_samples_per_second": 641.085,
-      "eval_steps_per_second": 5.029,
+      "eval_loss": 1.0985984802246094,
+      "eval_runtime": 18.4293,
+      "eval_samples_per_second": 532.575,
+      "eval_steps_per_second": 4.178,
       "step": 15340
     },
     {
       "epoch": 6.0,
-      "learning_rate": 0.000440019556714472,
-      "loss": 1.0986,
+      "learning_rate": 3.52e-05,
+      "loss": 1.0987,
       "step": 18408
     },
     {
       "epoch": 6.0,
       "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0983835458755493,
-      "eval_runtime": 15.3104,
-      "eval_samples_per_second": 641.066,
-      "eval_steps_per_second": 5.029,
+      "eval_loss": 1.098618745803833,
+      "eval_runtime": 18.4765,
+      "eval_samples_per_second": 531.214,
+      "eval_steps_per_second": 4.167,
       "step": 18408
     },
     {
       "epoch": 7.0,
-      "learning_rate": 0.0004305704041720991,
-      "loss": 1.0991,
+      "learning_rate": 3.44e-05,
+      "loss": 1.0986,
       "step": 21476
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": NaN,
-      "eval_runtime": 15.3384,
-      "eval_samples_per_second": 639.896,
-      "eval_steps_per_second": 5.02,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.098617434501648,
+      "eval_runtime": 18.4088,
+      "eval_samples_per_second": 533.17,
+      "eval_steps_per_second": 4.183,
       "step": 21476
     },
     {
       "epoch": 8.0,
-      "learning_rate": 0.00042057040417209905,
-      "loss": 1.1004,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.0986,
       "step": 24544
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": NaN,
-      "eval_runtime": 15.2542,
-      "eval_samples_per_second": 643.43,
-      "eval_steps_per_second": 5.048,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.0986160039901733,
+      "eval_runtime": 18.373,
+      "eval_samples_per_second": 534.207,
+      "eval_steps_per_second": 4.191,
       "step": 24544
     },
     {
-      "epoch": 9.0,
-      "learning_rate": 0.00041057040417209913,
-      "loss": 1.1004,
-      "step": 27612
-    },
-    {
-      "epoch": 9.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": NaN,
-      "eval_runtime": 15.3198,
-      "eval_samples_per_second": 640.675,
-      "eval_steps_per_second": 5.026,
-      "step": 27612
-    },
-    {
-      "epoch": 9.0,
-      "step": 27612,
-      "total_flos": 5.0478948198383616e+17,
-      "train_loss": 1.0999290425415127,
-      "train_runtime": 15619.8279,
-      "train_samples_per_second": 1257.063,
-      "train_steps_per_second": 9.821
+      "epoch": 8.0,
+      "step": 24544,
+      "total_flos": 4.487017617634099e+17,
+      "train_loss": 1.0992557968278907,
+      "train_runtime": 17750.054,
+      "train_samples_per_second": 1106.199,
+      "train_steps_per_second": 8.642
     }
   ],
   "max_steps": 153400,
   "num_train_epochs": 50,
-  "total_flos": 5.0478948198383616e+17,
+  "total_flos": 4.487017617634099e+17,
   "trial_name": null,
   "trial_params": null
 }