End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 tags:
 - generated_from_trainer
 datasets:
@@ -13,7 +15,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE QQP
       type: glue
       config: qqp
       split: validation
@@ -21,10 +23,10 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.8193173386099432
+      value: 0.7953747217412812
     - name: F1
       type: f1
-      value: 0.7510988449350915
+      value: 0.7269366603954186
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,12 +34,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hBERTv1_no_pretrain_qqp
 
-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 QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5245
-- Accuracy: 0.8193
-- F1: 0.7511
-- Combined Score: 0.7852
+- Loss: 0.4402
+- Accuracy: 0.7954
+- F1: 0.7269
+- Combined Score: 0.7612
 
 ## Model description
 

all_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 6.0,
-    "eval_accuracy": 0.6318327974276527,
-    "eval_combined_score": 0.3159163987138264,
-    "eval_f1": 0.0,
-    "eval_loss": 0.6598194241523743,
-    "eval_runtime": 70.3817,
+    "epoch": 8.0,
+    "eval_accuracy": 0.7953747217412812,
+    "eval_combined_score": 0.7611556910683499,
+    "eval_f1": 0.7269366603954186,
+    "eval_loss": 0.44018369913101196,
+    "eval_runtime": 188.2112,
     "eval_samples": 40430,
-    "eval_samples_per_second": 574.439,
-    "eval_steps_per_second": 4.49,
-    "train_loss": 0.4408971084521376,
-    "train_runtime": 11254.1141,
+    "eval_samples_per_second": 214.812,
+    "eval_steps_per_second": 2.242,
+    "train_loss": 0.3596843980101412,
+    "train_runtime": 42048.5829,
     "train_samples": 363846,
-    "train_samples_per_second": 1616.502,
-    "train_steps_per_second": 12.631
+    "train_samples_per_second": 432.65,
+    "train_steps_per_second": 4.508
 }

eval_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 6.0,
-    "eval_accuracy": 0.6318327974276527,
-    "eval_combined_score": 0.3159163987138264,
-    "eval_f1": 0.0,
-    "eval_loss": 0.6598194241523743,
-    "eval_runtime": 70.3817,
+    "epoch": 8.0,
+    "eval_accuracy": 0.7953747217412812,
+    "eval_combined_score": 0.7611556910683499,
+    "eval_f1": 0.7269366603954186,
+    "eval_loss": 0.44018369913101196,
+    "eval_runtime": 188.2112,
     "eval_samples": 40430,
-    "eval_samples_per_second": 574.439,
-    "eval_steps_per_second": 4.49
+    "eval_samples_per_second": 214.812,
+    "eval_steps_per_second": 2.242
 }

logs/events.out.tfevents.1686821332.garda.2667877.5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a619f558a948b495a2a8fdf5d3e07e6a0c0cb910cb8bf37f4d024af2046424de
+size 475

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 6.0,
-    "train_loss": 0.4408971084521376,
-    "train_runtime": 11254.1141,
+    "epoch": 8.0,
+    "train_loss": 0.3596843980101412,
+    "train_runtime": 42048.5829,
     "train_samples": 363846,
-    "train_samples_per_second": 1616.502,
-    "train_steps_per_second": 12.631
+    "train_samples_per_second": 432.65,
+    "train_steps_per_second": 4.508
 }

trainer_state.json

@@ -1,127 +1,161 @@
 {
-  "best_metric": 0.6598194241523743,
-  "best_model_checkpoint": "hBERTv1_no_pretrain_qqp/checkpoint-2843",
-  "epoch": 6.0,
-  "global_step": 17058,
+  "best_metric": 0.44018369913101196,
+  "best_model_checkpoint": "hBERTv1_no_pretrain_qqp/checkpoint-11373",
+  "epoch": 8.0,
+  "global_step": 30328,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "learning_rate": 0.0004900070348223708,
-      "loss": 0.6693,
-      "step": 2843
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.5334,
+      "step": 3791
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6318327974276527,
-      "eval_combined_score": 0.3159163987138264,
-      "eval_f1": 0.0,
-      "eval_loss": 0.6598194241523743,
-      "eval_runtime": 70.7024,
-      "eval_samples_per_second": 571.833,
-      "eval_steps_per_second": 4.469,
-      "step": 2843
+      "eval_accuracy": 0.7675735839722978,
+      "eval_combined_score": 0.7162763638068912,
+      "eval_f1": 0.6649791436414846,
+      "eval_loss": 0.4826056957244873,
+      "eval_runtime": 188.5117,
+      "eval_samples_per_second": 214.469,
+      "eval_steps_per_second": 2.239,
+      "step": 3791
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00048050650721069294,
-      "loss": 0.6591,
-      "step": 5686
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.4491,
+      "step": 7582
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.6318327974276527,
-      "eval_combined_score": 0.3159163987138264,
-      "eval_f1": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 70.3827,
-      "eval_samples_per_second": 574.431,
-      "eval_steps_per_second": 4.49,
-      "step": 5686
+      "eval_accuracy": 0.7908731140242394,
+      "eval_combined_score": 0.7417148741642232,
+      "eval_f1": 0.6925566343042071,
+      "eval_loss": 0.4492672383785248,
+      "eval_runtime": 188.5255,
+      "eval_samples_per_second": 214.454,
+      "eval_steps_per_second": 2.238,
+      "step": 7582
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00047059092507914177,
-      "loss": 0.6588,
-      "step": 8529
+      "learning_rate": 3.76e-05,
+      "loss": 0.3866,
+      "step": 11373
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.6318327974276527,
-      "eval_combined_score": 0.3159163987138264,
-      "eval_f1": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 70.6601,
-      "eval_samples_per_second": 572.176,
-      "eval_steps_per_second": 4.472,
-      "step": 8529
+      "eval_accuracy": 0.7953747217412812,
+      "eval_combined_score": 0.7611556910683499,
+      "eval_f1": 0.7269366603954186,
+      "eval_loss": 0.44018369913101196,
+      "eval_runtime": 188.9583,
+      "eval_samples_per_second": 213.963,
+      "eval_steps_per_second": 2.233,
+      "step": 11373
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00046059092507914174,
-      "loss": 0.6582,
-      "step": 11372
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.3657,
+      "step": 15164
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.6318327974276527,
-      "eval_combined_score": 0.3159163987138264,
-      "eval_f1": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 69.7585,
-      "eval_samples_per_second": 579.571,
-      "eval_steps_per_second": 4.53,
-      "step": 11372
+      "eval_accuracy": 0.777491961414791,
+      "eval_combined_score": 0.7493077018308378,
+      "eval_f1": 0.7211234422468845,
+      "eval_loss": 0.4990079700946808,
+      "eval_runtime": 188.3561,
+      "eval_samples_per_second": 214.647,
+      "eval_steps_per_second": 2.24,
+      "step": 15164
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.0004505909250791417,
-      "loss": 0.0,
-      "step": 14215
+      "learning_rate": 3.6e-05,
+      "loss": 0.3708,
+      "step": 18955
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.6318327974276527,
-      "eval_combined_score": 0.3159163987138264,
-      "eval_f1": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 69.723,
-      "eval_samples_per_second": 579.866,
-      "eval_steps_per_second": 4.532,
-      "step": 14215
+      "eval_accuracy": 0.8076675735839723,
+      "eval_combined_score": 0.7674797151079011,
+      "eval_f1": 0.72729185663183,
+      "eval_loss": 0.474396288394928,
+      "eval_runtime": 188.6326,
+      "eval_samples_per_second": 214.332,
+      "eval_steps_per_second": 2.237,
+      "step": 18955
     },
     {
       "epoch": 6.0,
-      "learning_rate": 0.0004405909250791418,
-      "loss": 0.0,
-      "step": 17058
+      "learning_rate": 3.52e-05,
+      "loss": 0.2948,
+      "step": 22746
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.6318327974276527,
-      "eval_combined_score": 0.3159163987138264,
-      "eval_f1": 0.0,
-      "eval_loss": NaN,
-      "eval_runtime": 69.5815,
-      "eval_samples_per_second": 581.045,
-      "eval_steps_per_second": 4.541,
-      "step": 17058
+      "eval_accuracy": 0.8143210487261934,
+      "eval_combined_score": 0.7760979872633377,
+      "eval_f1": 0.737874925800482,
+      "eval_loss": 0.4692937135696411,
+      "eval_runtime": 189.2189,
+      "eval_samples_per_second": 213.668,
+      "eval_steps_per_second": 2.23,
+      "step": 22746
     },
     {
-      "epoch": 6.0,
-      "step": 17058,
-      "total_flos": 3.2330651506468454e+17,
-      "train_loss": 0.4408971084521376,
-      "train_runtime": 11254.1141,
-      "train_samples_per_second": 1616.502,
-      "train_steps_per_second": 12.631
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.2546,
+      "step": 26537
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8119960425426663,
+      "eval_combined_score": 0.7848819882556219,
+      "eval_f1": 0.7577679339685777,
+      "eval_loss": 0.45072534680366516,
+      "eval_runtime": 189.1806,
+      "eval_samples_per_second": 213.711,
+      "eval_steps_per_second": 2.231,
+      "step": 26537
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.2225,
+      "step": 30328
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8193173386099432,
+      "eval_combined_score": 0.7852080917725173,
+      "eval_f1": 0.7510988449350915,
+      "eval_loss": 0.5244665145874023,
+      "eval_runtime": 189.2507,
+      "eval_samples_per_second": 213.632,
+      "eval_steps_per_second": 2.23,
+      "step": 30328
+    },
+    {
+      "epoch": 8.0,
+      "step": 30328,
+      "total_flos": 4.310753534195794e+17,
+      "train_loss": 0.3596843980101412,
+      "train_runtime": 42048.5829,
+      "train_samples_per_second": 432.65,
+      "train_steps_per_second": 4.508
     }
   ],
-  "max_steps": 142150,
+  "max_steps": 189550,
   "num_train_epochs": 50,
-  "total_flos": 3.2330651506468454e+17,
+  "total_flos": 4.310753534195794e+17,
   "trial_name": null,
   "trial_params": null
 }