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: 0.5773384587223137
+      value: 0.6009518579535054
 ---
 
 <!-- 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_48_KD_qnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_48_KD](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_48_KD) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_48_KD](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_48_KD) on the GLUE QNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6780
-- Accuracy: 0.5773
+- Loss: 0.6648
+- Accuracy: 0.6010
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 11.0,
+    "eval_accuracy": 0.6009518579535054,
+    "eval_loss": 0.6648141145706177,
+    "eval_runtime": 11.0952,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 492.376,
+    "eval_steps_per_second": 3.876,
+    "train_loss": 0.6673924017063666,
+    "train_runtime": 7878.8406,
+    "train_samples": 104743,
+    "train_samples_per_second": 664.711,
+    "train_steps_per_second": 5.197
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 11.0,
+    "eval_accuracy": 0.6009518579535054,
+    "eval_loss": 0.6648141145706177,
+    "eval_runtime": 11.0952,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 492.376,
+    "eval_steps_per_second": 3.876
+}

logs/events.out.tfevents.1686771642.serv-9207.3372160.7

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae6fffd5a307a264920a636540bb19f17c90ede84e8a0a4d848da9a83dc29171
+size 363

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 11.0,
+    "train_loss": 0.6673924017063666,
+    "train_runtime": 7878.8406,
+    "train_samples": 104743,
+    "train_samples_per_second": 664.711,
+    "train_steps_per_second": 5.197
+}

trainer_state.json

@@ -0,0 +1,190 @@
+{
+  "best_metric": 0.6648141145706177,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_48_KD_qnli/checkpoint-4914",
+  "epoch": 11.0,
+  "global_step": 9009,
+  "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": 0.6818,
+      "step": 819
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5965586673988651,
+      "eval_loss": 0.6669142246246338,
+      "eval_runtime": 11.3303,
+      "eval_samples_per_second": 482.16,
+      "eval_steps_per_second": 3.795,
+      "step": 819
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.6689,
+      "step": 1638
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5857587406187077,
+      "eval_loss": 0.6732200980186462,
+      "eval_runtime": 11.297,
+      "eval_samples_per_second": 483.58,
+      "eval_steps_per_second": 3.806,
+      "step": 1638
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.76e-05,
+      "loss": 0.6675,
+      "step": 2457
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5809994508511807,
+      "eval_loss": 0.672123908996582,
+      "eval_runtime": 11.2919,
+      "eval_samples_per_second": 483.798,
+      "eval_steps_per_second": 3.808,
+      "step": 2457
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.663,
+      "step": 3276
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5831960461285008,
+      "eval_loss": 0.6792891025543213,
+      "eval_runtime": 11.3369,
+      "eval_samples_per_second": 481.88,
+      "eval_steps_per_second": 3.793,
+      "step": 3276
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.66,
+      "step": 4095
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5998535603148453,
+      "eval_loss": 0.6663345098495483,
+      "eval_runtime": 11.3047,
+      "eval_samples_per_second": 483.249,
+      "eval_steps_per_second": 3.804,
+      "step": 4095
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.52e-05,
+      "loss": 0.6574,
+      "step": 4914
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6009518579535054,
+      "eval_loss": 0.6648141145706177,
+      "eval_runtime": 11.3263,
+      "eval_samples_per_second": 482.33,
+      "eval_steps_per_second": 3.796,
+      "step": 4914
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.6591,
+      "step": 5733
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5731283177741168,
+      "eval_loss": 0.6781439185142517,
+      "eval_runtime": 11.2984,
+      "eval_samples_per_second": 483.518,
+      "eval_steps_per_second": 3.806,
+      "step": 5733
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.659,
+      "step": 6552
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5950942705473183,
+      "eval_loss": 0.6685020923614502,
+      "eval_runtime": 11.3308,
+      "eval_samples_per_second": 482.137,
+      "eval_steps_per_second": 3.795,
+      "step": 6552
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.28e-05,
+      "loss": 0.6697,
+      "step": 7371
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5791689547867472,
+      "eval_loss": 0.6792623996734619,
+      "eval_runtime": 11.2936,
+      "eval_samples_per_second": 483.725,
+      "eval_steps_per_second": 3.807,
+      "step": 7371
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.6755,
+      "step": 8190
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.5698334248581366,
+      "eval_loss": 0.6829095482826233,
+      "eval_runtime": 11.3081,
+      "eval_samples_per_second": 483.107,
+      "eval_steps_per_second": 3.803,
+      "step": 8190
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.1200000000000006e-05,
+      "loss": 0.6794,
+      "step": 9009
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.5773384587223137,
+      "eval_loss": 0.6779787540435791,
+      "eval_runtime": 11.3367,
+      "eval_samples_per_second": 481.886,
+      "eval_steps_per_second": 3.793,
+      "step": 9009
+    },
+    {
+      "epoch": 11.0,
+      "step": 9009,
+      "total_flos": 1.706331010666332e+17,
+      "train_loss": 0.6673924017063666,
+      "train_runtime": 7878.8406,
+      "train_samples_per_second": 664.711,
+      "train_steps_per_second": 5.197
+    }
+  ],
+  "max_steps": 40950,
+  "num_train_epochs": 50,
+  "total_flos": 1.706331010666332e+17,
+  "trial_name": null,
+  "trial_params": null
+}