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 QNLI
       type: glue
       config: qnli
       split: validation
@@ -20,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.624016108365367
+      value: 0.6417719201903715
 ---
 
 <!-- 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_48_KD_w_init_qnli
 
-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 QNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8475
-- Accuracy: 0.6240
+- Loss: 0.6306
+- Accuracy: 0.6418
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.6417719201903715,
+    "eval_loss": 0.6305641531944275,
+    "eval_runtime": 11.4015,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 479.146,
+    "eval_steps_per_second": 3.771,
+    "train_loss": 0.5375828685590823,
+    "train_runtime": 4891.2045,
+    "train_samples": 104743,
+    "train_samples_per_second": 1070.728,
+    "train_steps_per_second": 8.372
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.6417719201903715,
+    "eval_loss": 0.6305641531944275,
+    "eval_runtime": 11.4015,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 479.146,
+    "eval_steps_per_second": 3.771
+}

logs/events.out.tfevents.1686870314.serv-9225.2647846.7

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30c11b0affb8cdf85592196418113b0b8150dc28588ba1ae3796c04531942c36
+size 363

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 7.0,
+    "train_loss": 0.5375828685590823,
+    "train_runtime": 4891.2045,
+    "train_samples": 104743,
+    "train_samples_per_second": 1070.728,
+    "train_steps_per_second": 8.372
+}

trainer_state.json

@@ -0,0 +1,130 @@
+{
+  "best_metric": 0.6305641531944275,
+  "best_model_checkpoint": "hBERTv2_new_pretrain_48_KD_w_init_qnli/checkpoint-1638",
+  "epoch": 7.0,
+  "global_step": 5733,
+  "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": 0.6733,
+      "step": 819
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5912502288120081,
+      "eval_loss": 0.6642831563949585,
+      "eval_runtime": 11.4659,
+      "eval_samples_per_second": 476.455,
+      "eval_steps_per_second": 3.75,
+      "step": 819
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.641,
+      "step": 1638
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6417719201903715,
+      "eval_loss": 0.6305641531944275,
+      "eval_runtime": 11.4569,
+      "eval_samples_per_second": 476.832,
+      "eval_steps_per_second": 3.753,
+      "step": 1638
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.76e-05,
+      "loss": 0.5952,
+      "step": 2457
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6377448288486179,
+      "eval_loss": 0.6487901210784912,
+      "eval_runtime": 11.4643,
+      "eval_samples_per_second": 476.521,
+      "eval_steps_per_second": 3.751,
+      "step": 2457
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.5439,
+      "step": 3276
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6302397949844408,
+      "eval_loss": 0.6660674810409546,
+      "eval_runtime": 11.4595,
+      "eval_samples_per_second": 476.722,
+      "eval_steps_per_second": 3.752,
+      "step": 3276
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.4907,
+      "step": 4095
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6252974556104705,
+      "eval_loss": 0.6936535239219666,
+      "eval_runtime": 11.4579,
+      "eval_samples_per_second": 476.79,
+      "eval_steps_per_second": 3.753,
+      "step": 4095
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.52e-05,
+      "loss": 0.4364,
+      "step": 4914
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6296906461651107,
+      "eval_loss": 0.7380545139312744,
+      "eval_runtime": 11.4626,
+      "eval_samples_per_second": 476.593,
+      "eval_steps_per_second": 3.751,
+      "step": 4914
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.3825,
+      "step": 5733
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.624016108365367,
+      "eval_loss": 0.8474876880645752,
+      "eval_runtime": 11.4645,
+      "eval_samples_per_second": 476.514,
+      "eval_steps_per_second": 3.751,
+      "step": 5733
+    },
+    {
+      "epoch": 7.0,
+      "step": 5733,
+      "total_flos": 1.0729602688011469e+17,
+      "train_loss": 0.5375828685590823,
+      "train_runtime": 4891.2045,
+      "train_samples_per_second": 1070.728,
+      "train_steps_per_second": 8.372
+    }
+  ],
+  "max_steps": 40950,
+  "num_train_epochs": 50,
+  "total_flos": 1.0729602688011469e+17,
+  "trial_name": null,
+  "trial_params": null
+}