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: 0.7147909967845659
+      value: 0.7335641850111303
     - name: F1
       type: f1
-      value: 0.5017069270990883
+      value: 0.6479278337037521
 ---
 
 <!-- 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. -->
 
 # hBERTv2_new_pretrain_48_KD_w_init_qqp
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new_48_KD_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new_48_KD_wt_init) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new_48_KD_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new_48_KD_wt_init) on the GLUE QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5929
-- Accuracy: 0.7148
-- F1: 0.5017
-- Combined Score: 0.6082
+- Loss: 0.5377
+- Accuracy: 0.7336
+- F1: 0.6479
+- Combined Score: 0.6907
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.7335641850111303,
+    "eval_combined_score": 0.6907460093574411,
+    "eval_f1": 0.6479278337037521,
+    "eval_loss": 0.5376537442207336,
+    "eval_runtime": 84.3159,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 479.506,
+    "eval_steps_per_second": 3.748,
+    "train_loss": 0.5469449936855403,
+    "train_runtime": 16965.2744,
+    "train_samples": 363846,
+    "train_samples_per_second": 1072.326,
+    "train_steps_per_second": 8.379
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.7335641850111303,
+    "eval_combined_score": 0.6907460093574411,
+    "eval_f1": 0.6479278337037521,
+    "eval_loss": 0.5376537442207336,
+    "eval_runtime": 84.3159,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 479.506,
+    "eval_steps_per_second": 3.748
+}

logs/events.out.tfevents.1686887680.serv-9225.2647846.9

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47d7a5dd8df6e6fa45932d86b01394d6183cc3e268ac034ae80b0e65646f66f5
+size 475

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 7.0,
+    "train_loss": 0.5469449936855403,
+    "train_runtime": 16965.2744,
+    "train_samples": 363846,
+    "train_samples_per_second": 1072.326,
+    "train_steps_per_second": 8.379
+}

trainer_state.json

@@ -0,0 +1,144 @@
+{
+  "best_metric": 0.5376537442207336,
+  "best_model_checkpoint": "hBERTv2_new_pretrain_48_KD_w_init_qqp/checkpoint-5686",
+  "epoch": 7.0,
+  "global_step": 19901,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.5751,
+      "step": 2843
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7259708137521642,
+      "eval_combined_score": 0.6813326769203516,
+      "eval_f1": 0.6366945400885391,
+      "eval_loss": 0.5381420254707336,
+      "eval_runtime": 84.8016,
+      "eval_samples_per_second": 476.76,
+      "eval_steps_per_second": 3.726,
+      "step": 2843
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.522,
+      "step": 5686
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7335641850111303,
+      "eval_combined_score": 0.6907460093574411,
+      "eval_f1": 0.6479278337037521,
+      "eval_loss": 0.5376537442207336,
+      "eval_runtime": 84.7312,
+      "eval_samples_per_second": 477.156,
+      "eval_steps_per_second": 3.729,
+      "step": 5686
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.76e-05,
+      "loss": 0.5274,
+      "step": 8529
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7332673757111056,
+      "eval_combined_score": 0.6739659254028979,
+      "eval_f1": 0.6146644750946902,
+      "eval_loss": 0.5656474828720093,
+      "eval_runtime": 84.8445,
+      "eval_samples_per_second": 476.519,
+      "eval_steps_per_second": 3.724,
+      "step": 8529
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.5435,
+      "step": 11372
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7424189957952015,
+      "eval_combined_score": 0.6868909138268045,
+      "eval_f1": 0.6313628318584072,
+      "eval_loss": 0.5560992956161499,
+      "eval_runtime": 84.7891,
+      "eval_samples_per_second": 476.83,
+      "eval_steps_per_second": 3.727,
+      "step": 11372
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.5405,
+      "step": 14215
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.7359139253029928,
+      "eval_combined_score": 0.6451016088631433,
+      "eval_f1": 0.5542892924232937,
+      "eval_loss": 0.5771732330322266,
+      "eval_runtime": 84.8467,
+      "eval_samples_per_second": 476.506,
+      "eval_steps_per_second": 3.724,
+      "step": 14215
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.52e-05,
+      "loss": 0.5468,
+      "step": 17058
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.7354687113529558,
+      "eval_combined_score": 0.6649394070625751,
+      "eval_f1": 0.5944101027721946,
+      "eval_loss": 0.5563399791717529,
+      "eval_runtime": 84.8147,
+      "eval_samples_per_second": 476.686,
+      "eval_steps_per_second": 3.726,
+      "step": 17058
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.5732,
+      "step": 19901
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7147909967845659,
+      "eval_combined_score": 0.6082489619418271,
+      "eval_f1": 0.5017069270990883,
+      "eval_loss": 0.5929098129272461,
+      "eval_runtime": 84.7761,
+      "eval_samples_per_second": 476.903,
+      "eval_steps_per_second": 3.727,
+      "step": 19901
+    },
+    {
+      "epoch": 7.0,
+      "step": 19901,
+      "total_flos": 3.727144401533665e+17,
+      "train_loss": 0.5469449936855403,
+      "train_runtime": 16965.2744,
+      "train_samples_per_second": 1072.326,
+      "train_steps_per_second": 8.379
+    }
+  ],
+  "max_steps": 142150,
+  "num_train_epochs": 50,
+  "total_flos": 3.727144401533665e+17,
+  "trial_name": null,
+  "trial_params": null
+}