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: 0.678921568627451
+      value: 0.6838235294117647
     - name: F1
       type: f1
-      value: 0.7827529021558872
+      value: 0.8122270742358079
 ---
 
 <!-- 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_w_init_48_mrpc
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48](https://huggingface.co/gokuls/bert_12_layer_model_v1_complete_training_new_wt_init_48) on the GLUE MRPC dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7737
-- Accuracy: 0.6789
-- F1: 0.7828
-- Combined Score: 0.7308
+- Loss: 0.6229
+- Accuracy: 0.6838
+- F1: 0.8122
+- Combined Score: 0.7480
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 20.0,
+    "eval_accuracy": 0.6838235294117647,
+    "eval_combined_score": 0.7480253018237863,
+    "eval_f1": 0.8122270742358079,
+    "eval_loss": 0.6228601932525635,
+    "eval_runtime": 0.7839,
+    "eval_samples": 408,
+    "eval_samples_per_second": 520.502,
+    "eval_steps_per_second": 5.103,
+    "train_loss": 0.6208314172152816,
+    "train_runtime": 514.9455,
+    "train_samples": 3668,
+    "train_samples_per_second": 356.154,
+    "train_steps_per_second": 2.816
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 20.0,
+    "eval_accuracy": 0.6838235294117647,
+    "eval_combined_score": 0.7480253018237863,
+    "eval_f1": 0.8122270742358079,
+    "eval_loss": 0.6228601932525635,
+    "eval_runtime": 0.7839,
+    "eval_samples": 408,
+    "eval_samples_per_second": 520.502,
+    "eval_steps_per_second": 5.103
+}

logs/events.out.tfevents.1686034971.serv-9207.2389173.8

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:033fffb5809606397f25123fa76e6379507174b7aff65c9e180fd71677f38288
+size 467

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 20.0,
+    "train_loss": 0.6208314172152816,
+    "train_runtime": 514.9455,
+    "train_samples": 3668,
+    "train_samples_per_second": 356.154,
+    "train_steps_per_second": 2.816
+}

trainer_state.json

@@ -0,0 +1,365 @@
+{
+  "best_metric": 0.6228601932525635,
+  "best_model_checkpoint": "hBERTv1_new_pretrain_w_init_48_mrpc/checkpoint-435",
+  "epoch": 20.0,
+  "global_step": 580,
+  "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": 0.6607,
+      "step": 29
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6262302994728088,
+      "eval_runtime": 0.7529,
+      "eval_samples_per_second": 541.921,
+      "eval_steps_per_second": 5.313,
+      "step": 29
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.6421,
+      "step": 58
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6368240714073181,
+      "eval_runtime": 0.7544,
+      "eval_samples_per_second": 540.804,
+      "eval_steps_per_second": 5.302,
+      "step": 58
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.76e-05,
+      "loss": 0.6411,
+      "step": 87
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6258090138435364,
+      "eval_runtime": 0.7525,
+      "eval_samples_per_second": 542.187,
+      "eval_steps_per_second": 5.316,
+      "step": 87
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 0.6406,
+      "step": 116
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.642214298248291,
+      "eval_runtime": 0.7556,
+      "eval_samples_per_second": 539.982,
+      "eval_steps_per_second": 5.294,
+      "step": 116
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.6364,
+      "step": 145
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6263068914413452,
+      "eval_runtime": 0.7566,
+      "eval_samples_per_second": 539.263,
+      "eval_steps_per_second": 5.287,
+      "step": 145
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.52e-05,
+      "loss": 0.6322,
+      "step": 174
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6252824664115906,
+      "eval_runtime": 0.7567,
+      "eval_samples_per_second": 539.166,
+      "eval_steps_per_second": 5.286,
+      "step": 174
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.44e-05,
+      "loss": 0.6398,
+      "step": 203
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6288966536521912,
+      "eval_runtime": 0.7572,
+      "eval_samples_per_second": 538.852,
+      "eval_steps_per_second": 5.283,
+      "step": 203
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.6363,
+      "step": 232
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6266850233078003,
+      "eval_runtime": 0.7574,
+      "eval_samples_per_second": 538.675,
+      "eval_steps_per_second": 5.281,
+      "step": 232
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.28e-05,
+      "loss": 0.6374,
+      "step": 261
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6375421285629272,
+      "eval_runtime": 0.7635,
+      "eval_samples_per_second": 534.369,
+      "eval_steps_per_second": 5.239,
+      "step": 261
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.6374,
+      "step": 290
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6247941851615906,
+      "eval_runtime": 0.7581,
+      "eval_samples_per_second": 538.207,
+      "eval_steps_per_second": 5.277,
+      "step": 290
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.1200000000000006e-05,
+      "loss": 0.638,
+      "step": 319
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.626215934753418,
+      "eval_runtime": 0.7576,
+      "eval_samples_per_second": 538.578,
+      "eval_steps_per_second": 5.28,
+      "step": 319
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 3.0400000000000004e-05,
+      "loss": 0.6353,
+      "step": 348
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6236165165901184,
+      "eval_runtime": 0.7585,
+      "eval_samples_per_second": 537.918,
+      "eval_steps_per_second": 5.274,
+      "step": 348
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 2.96e-05,
+      "loss": 0.6338,
+      "step": 377
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6263307929039001,
+      "eval_runtime": 0.7635,
+      "eval_samples_per_second": 534.401,
+      "eval_steps_per_second": 5.239,
+      "step": 377
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 2.8800000000000002e-05,
+      "loss": 0.637,
+      "step": 406
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6249856352806091,
+      "eval_runtime": 0.7626,
+      "eval_samples_per_second": 535.013,
+      "eval_steps_per_second": 5.245,
+      "step": 406
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 2.8e-05,
+      "loss": 0.6375,
+      "step": 435
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6228601932525635,
+      "eval_runtime": 0.7595,
+      "eval_samples_per_second": 537.205,
+      "eval_steps_per_second": 5.267,
+      "step": 435
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 0.7037,
+      "step": 464
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.7480253018237863,
+      "eval_f1": 0.8122270742358079,
+      "eval_loss": 0.6437533497810364,
+      "eval_runtime": 0.76,
+      "eval_samples_per_second": 536.826,
+      "eval_steps_per_second": 5.263,
+      "step": 464
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 2.6400000000000005e-05,
+      "loss": 0.6198,
+      "step": 493
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.696078431372549,
+      "eval_combined_score": 0.7499379498634897,
+      "eval_f1": 0.8037974683544306,
+      "eval_loss": 0.6241622567176819,
+      "eval_runtime": 0.7598,
+      "eval_samples_per_second": 536.975,
+      "eval_steps_per_second": 5.264,
+      "step": 493
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 0.5847,
+      "step": 522
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.6740196078431373,
+      "eval_combined_score": 0.724106578115117,
+      "eval_f1": 0.7741935483870968,
+      "eval_loss": 0.625978946685791,
+      "eval_runtime": 0.7574,
+      "eval_samples_per_second": 538.703,
+      "eval_steps_per_second": 5.281,
+      "step": 522
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 2.4800000000000003e-05,
+      "loss": 0.4983,
+      "step": 551
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.7034313725490197,
+      "eval_combined_score": 0.7596304503536575,
+      "eval_f1": 0.8158295281582952,
+      "eval_loss": 0.717353343963623,
+      "eval_runtime": 0.7597,
+      "eval_samples_per_second": 537.047,
+      "eval_steps_per_second": 5.265,
+      "step": 551
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 2.4e-05,
+      "loss": 0.4245,
+      "step": 580
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.678921568627451,
+      "eval_combined_score": 0.7308372353916691,
+      "eval_f1": 0.7827529021558872,
+      "eval_loss": 0.7737127542495728,
+      "eval_runtime": 0.7589,
+      "eval_samples_per_second": 537.61,
+      "eval_steps_per_second": 5.271,
+      "step": 580
+    },
+    {
+      "epoch": 20.0,
+      "step": 580,
+      "total_flos": 1.086437948653568e+16,
+      "train_loss": 0.6208314172152816,
+      "train_runtime": 514.9455,
+      "train_samples_per_second": 356.154,
+      "train_steps_per_second": 2.816
+    }
+  ],
+  "max_steps": 1450,
+  "num_train_epochs": 50,
+  "total_flos": 1.086437948653568e+16,
+  "trial_name": null,
+  "trial_params": null
+}