End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 base_model: gokuls/bert_12_layer_model_v2_complete_training_new_48
 tags:
 - generated_from_trainer
@@ -13,7 +15,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE QNLI
       type: glue
       config: qnli
       split: validation
@@ -21,7 +23,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.5328574043565807
+      value: 0.5839282445542742
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -29,10 +31,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hBERTv2_new_pretrain_48_ver2_qnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new_48](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new_48) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v2_complete_training_new_48](https://huggingface.co/gokuls/bert_12_layer_model_v2_complete_training_new_48) on the GLUE QNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6932
-- Accuracy: 0.5329
+- Loss: 0.6731
+- Accuracy: 0.5839
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5839282445542742,
+    "eval_loss": 0.6731342077255249,
+    "eval_runtime": 27.9584,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 195.398,
+    "eval_steps_per_second": 3.076,
+    "train_loss": 0.6740620123153062,
+    "train_runtime": 13295.2946,
+    "train_samples": 104743,
+    "train_samples_per_second": 118.173,
+    "train_steps_per_second": 1.847
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5839282445542742,
+    "eval_loss": 0.6731342077255249,
+    "eval_runtime": 27.9584,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 195.398,
+    "eval_steps_per_second": 3.076
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "train_loss": 0.6740620123153062,
+    "train_runtime": 13295.2946,
+    "train_samples": 104743,
+    "train_samples_per_second": 118.173,
+    "train_steps_per_second": 1.847
+}

trainer_state.json

@@ -0,0 +1,148 @@
+{
+  "best_metric": 0.6731342077255249,
+  "best_model_checkpoint": "hBERTv2_new_pretrain_48_ver2_qnli/checkpoint-4911",
+  "epoch": 8.0,
+  "eval_steps": 500,
+  "global_step": 13096,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 0.6905,
+      "step": 1637
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5723961193483434,
+      "eval_loss": 0.6829714179039001,
+      "eval_runtime": 28.5405,
+      "eval_samples_per_second": 191.413,
+      "eval_steps_per_second": 3.013,
+      "step": 1637
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.466666666666667e-05,
+      "loss": 0.6788,
+      "step": 3274
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.581182500457624,
+      "eval_loss": 0.6778351664543152,
+      "eval_runtime": 28.5366,
+      "eval_samples_per_second": 191.439,
+      "eval_steps_per_second": 3.014,
+      "step": 3274
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.6699,
+      "step": 4911
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5839282445542742,
+      "eval_loss": 0.6731342077255249,
+      "eval_runtime": 28.5826,
+      "eval_samples_per_second": 191.13,
+      "eval_steps_per_second": 3.009,
+      "step": 4911
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 2.9333333333333333e-05,
+      "loss": 0.6711,
+      "step": 6548
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5678198791872597,
+      "eval_loss": 0.6835048794746399,
+      "eval_runtime": 28.4908,
+      "eval_samples_per_second": 191.746,
+      "eval_steps_per_second": 3.019,
+      "step": 6548
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6723,
+      "step": 8185
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5546403075233388,
+      "eval_loss": 0.6851611733436584,
+      "eval_runtime": 28.4432,
+      "eval_samples_per_second": 192.067,
+      "eval_steps_per_second": 3.024,
+      "step": 8185
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.4e-05,
+      "loss": 0.674,
+      "step": 9822
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5678198791872597,
+      "eval_loss": 0.6817928552627563,
+      "eval_runtime": 28.4366,
+      "eval_samples_per_second": 192.112,
+      "eval_steps_per_second": 3.024,
+      "step": 9822
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 0.6679,
+      "step": 11459
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5736774665934469,
+      "eval_loss": 0.6758979558944702,
+      "eval_runtime": 28.446,
+      "eval_samples_per_second": 192.048,
+      "eval_steps_per_second": 3.023,
+      "step": 11459
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.6681,
+      "step": 13096
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5328574043565807,
+      "eval_loss": 0.6932294368743896,
+      "eval_runtime": 28.408,
+      "eval_samples_per_second": 192.305,
+      "eval_steps_per_second": 3.027,
+      "step": 13096
+    },
+    {
+      "epoch": 8.0,
+      "step": 13096,
+      "total_flos": 1.2262403072013107e+17,
+      "train_loss": 0.6740620123153062,
+      "train_runtime": 13295.2946,
+      "train_samples_per_second": 118.173,
+      "train_steps_per_second": 1.847
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 24555,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "total_flos": 1.2262403072013107e+17,
+  "trial_name": null,
+  "trial_params": null
+}