End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 tags:
 - generated_from_trainer
 datasets:
@@ -13,7 +15,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE MRPC
       type: glue
       config: mrpc
       split: validation
@@ -21,10 +23,10 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.6838235294117647
+      value: 0.7156862745098039
     - name: F1
       type: f1
-      value: 0.7794871794871795
+      value: 0.8104575163398692
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,12 +34,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 # hBERTv1_new_pretrain_48_KD_w_init_mrpc
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_48_KD_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_48_KD_wt_init) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_48_KD_wt_init](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_48_KD_wt_init) on the GLUE MRPC dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7273
-- Accuracy: 0.6838
-- F1: 0.7795
-- Combined Score: 0.7317
+- Loss: 0.5878
+- Accuracy: 0.7157
+- F1: 0.8105
+- Combined Score: 0.7631
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.7156862745098039,
+    "eval_combined_score": 0.7630718954248366,
+    "eval_f1": 0.8104575163398692,
+    "eval_loss": 0.5877546668052673,
+    "eval_runtime": 0.895,
+    "eval_samples": 408,
+    "eval_samples_per_second": 455.868,
+    "eval_steps_per_second": 4.469,
+    "train_loss": 0.5418458026030968,
+    "train_runtime": 262.0538,
+    "train_samples": 3668,
+    "train_samples_per_second": 699.856,
+    "train_steps_per_second": 5.533
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.7156862745098039,
+    "eval_combined_score": 0.7630718954248366,
+    "eval_f1": 0.8104575163398692,
+    "eval_loss": 0.5877546668052673,
+    "eval_runtime": 0.895,
+    "eval_samples": 408,
+    "eval_samples_per_second": 455.868,
+    "eval_steps_per_second": 4.469
+}

logs/events.out.tfevents.1686865347.serv-3317.1588421.5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7093d7b4f09fd45233cf0a6ca4f3dfaaec2009fd4a73548492a3976089a245a4
+size 467

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "train_loss": 0.5418458026030968,
+    "train_runtime": 262.0538,
+    "train_samples": 3668,
+    "train_samples_per_second": 699.856,
+    "train_steps_per_second": 5.533
+}

trainer_state.json

@@ -0,0 +1,161 @@
+{
+  "best_metric": 0.5877546668052673,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_48_KD_w_init_mrpc/checkpoint-87",
+  "epoch": 8.0,
+  "global_step": 232,
+  "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": 0.6514,
+      "step": 29
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.6887254901960784,
+      "eval_combined_score": 0.7516620151710319,
+      "eval_f1": 0.8145985401459854,
+      "eval_loss": 0.6205276250839233,
+      "eval_runtime": 0.8971,
+      "eval_samples_per_second": 454.795,
+      "eval_steps_per_second": 4.459,
+      "step": 29
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.619,
+      "step": 58
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6617647058823529,
+      "eval_combined_score": 0.6992029636281993,
+      "eval_f1": 0.7366412213740458,
+      "eval_loss": 0.6164662837982178,
+      "eval_runtime": 0.9039,
+      "eval_samples_per_second": 451.393,
+      "eval_steps_per_second": 4.425,
+      "step": 58
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.76e-05,
+      "loss": 0.6208,
+      "step": 87
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7156862745098039,
+      "eval_combined_score": 0.7630718954248366,
+      "eval_f1": 0.8104575163398692,
+      "eval_loss": 0.5877546668052673,
+      "eval_runtime": 0.8941,
+      "eval_samples_per_second": 456.301,
+      "eval_steps_per_second": 4.474,
+      "step": 87
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.578,
+      "step": 116
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7132352941176471,
+      "eval_combined_score": 0.7559291789004758,
+      "eval_f1": 0.7986230636833047,
+      "eval_loss": 0.595228374004364,
+      "eval_runtime": 0.8967,
+      "eval_samples_per_second": 455.006,
+      "eval_steps_per_second": 4.461,
+      "step": 116
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.5612,
+      "step": 145
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6936274509803921,
+      "eval_combined_score": 0.7417717086834734,
+      "eval_f1": 0.7899159663865546,
+      "eval_loss": 0.5909532904624939,
+      "eval_runtime": 0.9008,
+      "eval_samples_per_second": 452.945,
+      "eval_steps_per_second": 4.441,
+      "step": 145
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.52e-05,
+      "loss": 0.4844,
+      "step": 174
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6519607843137255,
+      "eval_combined_score": 0.6904842089507559,
+      "eval_f1": 0.7290076335877863,
+      "eval_loss": 0.6260918974876404,
+      "eval_runtime": 0.9039,
+      "eval_samples_per_second": 451.357,
+      "eval_steps_per_second": 4.425,
+      "step": 174
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.4281,
+      "step": 203
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7009803921568627,
+      "eval_combined_score": 0.7471003655699567,
+      "eval_f1": 0.7932203389830508,
+      "eval_loss": 0.6146315932273865,
+      "eval_runtime": 0.9,
+      "eval_samples_per_second": 453.349,
+      "eval_steps_per_second": 4.445,
+      "step": 203
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.3919,
+      "step": 232
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7316553544494722,
+      "eval_f1": 0.7794871794871795,
+      "eval_loss": 0.7272526025772095,
+      "eval_runtime": 0.8993,
+      "eval_samples_per_second": 453.667,
+      "eval_steps_per_second": 4.448,
+      "step": 232
+    },
+    {
+      "epoch": 8.0,
+      "step": 232,
+      "total_flos": 4345751794614272.0,
+      "train_loss": 0.5418458026030968,
+      "train_runtime": 262.0538,
+      "train_samples_per_second": 699.856,
+      "train_steps_per_second": 5.533
+    }
+  ],
+  "max_steps": 1450,
+  "num_train_epochs": 50,
+  "total_flos": 4345751794614272.0,
+  "trial_name": null,
+  "trial_params": null
+}