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: 0.5824638477027274
+      value: 0.5983891634632985
 ---
 
 <!-- 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_emb_com_qnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48) on the GLUE QNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7175
-- Accuracy: 0.5825
+- Loss: 0.6611
+- Accuracy: 0.5984
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5983891634632985,
+    "eval_loss": 0.6611328721046448,
+    "eval_runtime": 24.4827,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 223.137,
+    "eval_steps_per_second": 1.756,
+    "train_loss": 0.6348617563026437,
+    "train_runtime": 11786.3018,
+    "train_samples": 104743,
+    "train_samples_per_second": 444.342,
+    "train_steps_per_second": 3.474
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5983891634632985,
+    "eval_loss": 0.6611328721046448,
+    "eval_runtime": 24.4827,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 223.137,
+    "eval_steps_per_second": 1.756
+}

logs/events.out.tfevents.1686771397.garda.2601621.7

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd653c6ab1b2c02373961930ffa77f223faa33936cad8d6a8dadf4e9f66e2f16
+size 363

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "train_loss": 0.6348617563026437,
+    "train_runtime": 11786.3018,
+    "train_samples": 104743,
+    "train_samples_per_second": 444.342,
+    "train_steps_per_second": 3.474
+}

trainer_state.json

@@ -0,0 +1,145 @@
+{
+  "best_metric": 0.6611328721046448,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_48_emb_com_qnli/checkpoint-2457",
+  "epoch": 8.0,
+  "global_step": 6552,
+  "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": 0.6862,
+      "step": 819
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5513454146073586,
+      "eval_loss": 0.6995250582695007,
+      "eval_runtime": 24.6277,
+      "eval_samples_per_second": 221.823,
+      "eval_steps_per_second": 1.746,
+      "step": 819
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.6677,
+      "step": 1638
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.594911220940875,
+      "eval_loss": 0.6624441146850586,
+      "eval_runtime": 24.5823,
+      "eval_samples_per_second": 222.233,
+      "eval_steps_per_second": 1.749,
+      "step": 1638
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.76e-05,
+      "loss": 0.6553,
+      "step": 2457
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5983891634632985,
+      "eval_loss": 0.6611328721046448,
+      "eval_runtime": 24.601,
+      "eval_samples_per_second": 222.064,
+      "eval_steps_per_second": 1.748,
+      "step": 2457
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.6426,
+      "step": 3276
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5762401610836537,
+      "eval_loss": 0.6866379976272583,
+      "eval_runtime": 24.809,
+      "eval_samples_per_second": 220.203,
+      "eval_steps_per_second": 1.733,
+      "step": 3276
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.6244,
+      "step": 4095
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5950942705473183,
+      "eval_loss": 0.675081729888916,
+      "eval_runtime": 24.7609,
+      "eval_samples_per_second": 220.63,
+      "eval_steps_per_second": 1.737,
+      "step": 4095
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.52e-05,
+      "loss": 0.6136,
+      "step": 4914
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5727622185612301,
+      "eval_loss": 0.6902745962142944,
+      "eval_runtime": 24.6831,
+      "eval_samples_per_second": 221.326,
+      "eval_steps_per_second": 1.742,
+      "step": 4914
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.6036,
+      "step": 5733
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5731283177741168,
+      "eval_loss": 0.7188088297843933,
+      "eval_runtime": 24.6865,
+      "eval_samples_per_second": 221.295,
+      "eval_steps_per_second": 1.742,
+      "step": 5733
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.5855,
+      "step": 6552
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5824638477027274,
+      "eval_loss": 0.7175068259239197,
+      "eval_runtime": 24.6261,
+      "eval_samples_per_second": 221.838,
+      "eval_steps_per_second": 1.746,
+      "step": 6552
+    },
+    {
+      "epoch": 8.0,
+      "step": 6552,
+      "total_flos": 1.2422266908730982e+17,
+      "train_loss": 0.6348617563026437,
+      "train_runtime": 11786.3018,
+      "train_samples_per_second": 444.342,
+      "train_steps_per_second": 3.474
+    }
+  ],
+  "max_steps": 40950,
+  "num_train_epochs": 50,
+  "total_flos": 1.2422266908730982e+17,
+  "trial_name": null,
+  "trial_params": null
+}