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 MNLI
       type: glue
       config: mnli
       split: validation_matched
@@ -20,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.3544574630667346
+      value: 0.3295362082994304
 ---
 
 <!-- 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_mnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: nan
-- Accuracy: 0.3545
+- Loss: 15.1649
+- Accuracy: 0.3295
 
 ## Model description
 

all_results.json

@@ -0,0 +1,21 @@
+{
+    "epoch": 6.0,
+    "epoch_mm": 6.0,
+    "eval_accuracy": 0.3273560876209883,
+    "eval_accuracy_mm": 0.3295362082994304,
+    "eval_loss": 15.246305465698242,
+    "eval_loss_mm": 15.164895057678223,
+    "eval_runtime": 17.7885,
+    "eval_runtime_mm": 17.8247,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 551.762,
+    "eval_samples_per_second_mm": 551.595,
+    "eval_steps_per_second": 4.329,
+    "eval_steps_per_second_mm": 4.32,
+    "train_loss": 24.543312944779444,
+    "train_runtime": 12385.4858,
+    "train_samples": 392702,
+    "train_samples_per_second": 1585.331,
+    "train_steps_per_second": 12.385
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 6.0,
+    "epoch_mm": 6.0,
+    "eval_accuracy": 0.3273560876209883,
+    "eval_accuracy_mm": 0.3295362082994304,
+    "eval_loss": 15.246305465698242,
+    "eval_loss_mm": 15.164895057678223,
+    "eval_runtime": 17.7885,
+    "eval_runtime_mm": 17.8247,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 551.762,
+    "eval_samples_per_second_mm": 551.595,
+    "eval_steps_per_second": 4.329,
+    "eval_steps_per_second_mm": 4.32
+}

logs/events.out.tfevents.1685560589.serv-3318.2597307.26

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e9a423ada71411c1505931d6016bdbc303a2e3f7370575fc215f329ba7ffb81
+size 698

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 6.0,
+    "train_loss": 24.543312944779444,
+    "train_runtime": 12385.4858,
+    "train_samples": 392702,
+    "train_samples_per_second": 1585.331,
+    "train_steps_per_second": 12.385
+}

trainer_state.json

@@ -0,0 +1,115 @@
+{
+  "best_metric": 15.246305465698242,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_mnli/checkpoint-3068",
+  "epoch": 6.0,
+  "global_step": 18408,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00049,
+      "loss": 30.2532,
+      "step": 3068
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 15.246305465698242,
+      "eval_runtime": 17.8319,
+      "eval_samples_per_second": 550.419,
+      "eval_steps_per_second": 4.318,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00048,
+      "loss": 28.5719,
+      "step": 6136
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.3521141110545084,
+      "eval_loss": 22.4074764251709,
+      "eval_runtime": 17.8388,
+      "eval_samples_per_second": 550.205,
+      "eval_steps_per_second": 4.316,
+      "step": 6136
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.00047,
+      "loss": 29.4656,
+      "step": 9204
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.3521141110545084,
+      "eval_loss": 22.4074764251709,
+      "eval_runtime": 17.8305,
+      "eval_samples_per_second": 550.46,
+      "eval_steps_per_second": 4.318,
+      "step": 9204
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.00046,
+      "loss": 29.488,
+      "step": 12272
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.3521141110545084,
+      "eval_loss": 22.4074764251709,
+      "eval_runtime": 17.8149,
+      "eval_samples_per_second": 550.945,
+      "eval_steps_per_second": 4.322,
+      "step": 12272
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.00045000000000000004,
+      "loss": 29.4812,
+      "step": 15340
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": NaN,
+      "eval_runtime": 17.6387,
+      "eval_samples_per_second": 556.446,
+      "eval_steps_per_second": 4.365,
+      "step": 15340
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.00044,
+      "loss": 0.0,
+      "step": 18408
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": NaN,
+      "eval_runtime": 17.6186,
+      "eval_samples_per_second": 557.082,
+      "eval_steps_per_second": 4.37,
+      "step": 18408
+    },
+    {
+      "epoch": 6.0,
+      "step": 18408,
+      "total_flos": 3.48950175651201e+17,
+      "train_loss": 24.543312944779444,
+      "train_runtime": 12385.4858,
+      "train_samples_per_second": 1585.331,
+      "train_steps_per_second": 12.385
+    }
+  ],
+  "max_steps": 153400,
+  "num_train_epochs": 50,
+  "total_flos": 3.48950175651201e+17,
+  "trial_name": null,
+  "trial_params": null
+}