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: 0.31818644931227713
+      value: 0.3522172497965826
 ---
 
 <!-- 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_mnli
 
-This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1](https://huggingface.co/gokuls/bert_12_layer_model_v1) on the glue dataset.
+This model is a fine-tuned version of [gokuls/bert_12_layer_model_v1](https://huggingface.co/gokuls/bert_12_layer_model_v1) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0983
-- Accuracy: 0.3182
+- Loss: 1.0982
+- Accuracy: 0.3522
 
 ## Model description
 

all_results.json

@@ -0,0 +1,21 @@
+{
+    "epoch": 18.0,
+    "epoch_mm": 18.0,
+    "eval_accuracy": 0.3544574630667346,
+    "eval_accuracy_mm": 0.3522172497965826,
+    "eval_loss": 1.098163366317749,
+    "eval_loss_mm": 1.0981905460357666,
+    "eval_runtime": 12.6046,
+    "eval_runtime_mm": 12.5657,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 778.686,
+    "eval_samples_per_second_mm": 782.447,
+    "eval_steps_per_second": 3.094,
+    "eval_steps_per_second_mm": 3.104,
+    "train_loss": 1.0994189529164742,
+    "train_runtime": 25959.7521,
+    "train_samples": 392702,
+    "train_samples_per_second": 756.367,
+    "train_steps_per_second": 2.955
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 18.0,
+    "epoch_mm": 18.0,
+    "eval_accuracy": 0.3544574630667346,
+    "eval_accuracy_mm": 0.3522172497965826,
+    "eval_loss": 1.098163366317749,
+    "eval_loss_mm": 1.0981905460357666,
+    "eval_runtime": 12.6046,
+    "eval_runtime_mm": 12.5657,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 778.686,
+    "eval_samples_per_second_mm": 782.447,
+    "eval_steps_per_second": 3.094,
+    "eval_steps_per_second_mm": 3.104
+}

logs/events.out.tfevents.1677782974.serv-9223.249682.26

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c39512d49dbbd185467850846ba5a9e8ab955dac3a9daecef8eee6c0a7bfb525
+size 698

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 18.0,
+    "train_loss": 1.0994189529164742,
+    "train_runtime": 25959.7521,
+    "train_samples": 392702,
+    "train_samples_per_second": 756.367,
+    "train_steps_per_second": 2.955
+}

trainer_state.json

@@ -0,0 +1,295 @@
+{
+  "best_metric": 1.098163366317749,
+  "best_model_checkpoint": "hBERTv1_mnli/checkpoint-19942",
+  "epoch": 18.0,
+  "global_step": 27612,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.9e-05,
+      "loss": 1.1001,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0993714332580566,
+      "eval_runtime": 12.6071,
+      "eval_samples_per_second": 778.528,
+      "eval_steps_per_second": 3.093,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 1.0988,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0989975929260254,
+      "eval_runtime": 12.5819,
+      "eval_samples_per_second": 780.088,
+      "eval_steps_per_second": 3.1,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 1.0987,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.099189043045044,
+      "eval_runtime": 12.6822,
+      "eval_samples_per_second": 773.921,
+      "eval_steps_per_second": 3.075,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.0987,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.0986238718032837,
+      "eval_runtime": 12.5133,
+      "eval_samples_per_second": 784.368,
+      "eval_steps_per_second": 3.117,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.50006518904824e-05,
+      "loss": 1.0987,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.0984646081924438,
+      "eval_runtime": 12.6008,
+      "eval_samples_per_second": 778.92,
+      "eval_steps_per_second": 3.095,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.40013037809648e-05,
+      "loss": 1.0986,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.0987476110458374,
+      "eval_runtime": 12.4519,
+      "eval_samples_per_second": 788.232,
+      "eval_steps_per_second": 3.132,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3003259452411996e-05,
+      "loss": 1.105,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.0986319780349731,
+      "eval_runtime": 12.6436,
+      "eval_samples_per_second": 776.285,
+      "eval_steps_per_second": 3.085,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2003259452412e-05,
+      "loss": 1.1045,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0985896587371826,
+      "eval_runtime": 12.6272,
+      "eval_samples_per_second": 777.291,
+      "eval_steps_per_second": 3.089,
+      "step": 12272
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.1003259452411995e-05,
+      "loss": 1.0988,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.3273560876209883,
+      "eval_loss": 1.0983130931854248,
+      "eval_runtime": 12.4879,
+      "eval_samples_per_second": 785.962,
+      "eval_steps_per_second": 3.123,
+      "step": 13806
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4.0003259452412e-05,
+      "loss": 1.0987,
+      "step": 15340
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.098677158355713,
+      "eval_runtime": 12.4301,
+      "eval_samples_per_second": 789.616,
+      "eval_steps_per_second": 3.138,
+      "step": 15340
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.9003259452411994e-05,
+      "loss": 1.0987,
+      "step": 16874
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0991103649139404,
+      "eval_runtime": 12.5586,
+      "eval_samples_per_second": 781.533,
+      "eval_steps_per_second": 3.105,
+      "step": 16874
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 3.8003911342894394e-05,
+      "loss": 1.0986,
+      "step": 18408
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.0986063480377197,
+      "eval_runtime": 12.5872,
+      "eval_samples_per_second": 779.762,
+      "eval_steps_per_second": 3.098,
+      "step": 18408
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 3.700391134289439e-05,
+      "loss": 1.0986,
+      "step": 19942
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.098163366317749,
+      "eval_runtime": 12.4723,
+      "eval_samples_per_second": 786.944,
+      "eval_steps_per_second": 3.127,
+      "step": 19942
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.600456323337679e-05,
+      "loss": 1.0986,
+      "step": 21476
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.0988693237304688,
+      "eval_runtime": 12.5684,
+      "eval_samples_per_second": 780.924,
+      "eval_steps_per_second": 3.103,
+      "step": 21476
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.500521512385919e-05,
+      "loss": 1.0986,
+      "step": 23010
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.098739504814148,
+      "eval_runtime": 12.6125,
+      "eval_samples_per_second": 778.196,
+      "eval_steps_per_second": 3.092,
+      "step": 23010
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.4005215123859194e-05,
+      "loss": 1.0986,
+      "step": 24544
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.0985974073410034,
+      "eval_runtime": 12.4698,
+      "eval_samples_per_second": 787.103,
+      "eval_steps_per_second": 3.128,
+      "step": 24544
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3.300586701434159e-05,
+      "loss": 1.0986,
+      "step": 26078
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.3544574630667346,
+      "eval_loss": 1.098615288734436,
+      "eval_runtime": 12.5264,
+      "eval_samples_per_second": 783.545,
+      "eval_steps_per_second": 3.113,
+      "step": 26078
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 3.200586701434159e-05,
+      "loss": 1.0986,
+      "step": 27612
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.31818644931227713,
+      "eval_loss": 1.0982955694198608,
+      "eval_runtime": 12.8565,
+      "eval_samples_per_second": 763.425,
+      "eval_steps_per_second": 3.033,
+      "step": 27612
+    },
+    {
+      "epoch": 18.0,
+      "step": 27612,
+      "total_flos": 8.928550500811407e+17,
+      "train_loss": 1.0994189529164742,
+      "train_runtime": 25959.7521,
+      "train_samples_per_second": 756.367,
+      "train_steps_per_second": 2.955
+    }
+  ],
+  "max_steps": 76700,
+  "num_train_epochs": 50,
+  "total_flos": 8.928550500811407e+17,
+  "trial_name": null,
+  "trial_params": null
+}