End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 license: apache-2.0
 tags:
 - generated_from_trainer
@@ -13,7 +15,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE QNLI
       type: glue
       config: qnli
       split: validation
@@ -21,7 +23,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.5736774665934469
+      value: 0.5874061870766978
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -29,10 +31,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # distilbert_add_GLUE_Experiment_logit_kd_qnli_256
 
-This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the glue dataset.
+This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the GLUE QNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4100
-- Accuracy: 0.5737
+- Loss: 0.3989
+- Accuracy: 0.5874
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5874061870766978,
+    "eval_loss": 0.39891308546066284,
+    "eval_runtime": 5.3419,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 1022.662,
+    "eval_steps_per_second": 4.118,
+    "train_loss": 0.39033554356272626,
+    "train_runtime": 1120.6278,
+    "train_samples": 104743,
+    "train_samples_per_second": 4673.407,
+    "train_steps_per_second": 18.293
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5874061870766978,
+    "eval_loss": 0.39891308546066284,
+    "eval_runtime": 5.3419,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 1022.662,
+    "eval_steps_per_second": 4.118
+}

logs/events.out.tfevents.1674949063.serv-3334.530723.8

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef354c3e77c7d4ef8dc14a7ab6107a81d6c1d252df260001c71faaeeab4454c6
+size 363

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "train_loss": 0.39033554356272626,
+    "train_runtime": 1120.6278,
+    "train_samples": 104743,
+    "train_samples_per_second": 4673.407,
+    "train_steps_per_second": 18.293
+}

trainer_state.json

@@ -0,0 +1,145 @@
+{
+  "best_metric": 0.39891308546066284,
+  "best_model_checkpoint": "distilbert_add_GLUE_Experiment_logit_kd_qnli_256/checkpoint-1230",
+  "epoch": 8.0,
+  "global_step": 3280,
+  "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": 0.4156,
+      "step": 410
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5053999633900788,
+      "eval_loss": 0.4111415147781372,
+      "eval_runtime": 5.3768,
+      "eval_samples_per_second": 1016.031,
+      "eval_steps_per_second": 4.092,
+      "step": 410
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.4078,
+      "step": 820
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5799011532125206,
+      "eval_loss": 0.4018184542655945,
+      "eval_runtime": 5.3514,
+      "eval_samples_per_second": 1020.846,
+      "eval_steps_per_second": 4.111,
+      "step": 820
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3962,
+      "step": 1230
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5874061870766978,
+      "eval_loss": 0.39891308546066284,
+      "eval_runtime": 5.3383,
+      "eval_samples_per_second": 1023.358,
+      "eval_steps_per_second": 4.121,
+      "step": 1230
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.3899,
+      "step": 1640
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5866739886509243,
+      "eval_loss": 0.40177708864212036,
+      "eval_runtime": 5.3395,
+      "eval_samples_per_second": 1023.138,
+      "eval_steps_per_second": 4.12,
+      "step": 1640
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.3851,
+      "step": 2050
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5799011532125206,
+      "eval_loss": 0.4032135009765625,
+      "eval_runtime": 5.3369,
+      "eval_samples_per_second": 1023.624,
+      "eval_steps_per_second": 4.122,
+      "step": 2050
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.3802,
+      "step": 2460
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5727622185612301,
+      "eval_loss": 0.4117971658706665,
+      "eval_runtime": 5.4272,
+      "eval_samples_per_second": 1006.599,
+      "eval_steps_per_second": 4.054,
+      "step": 2460
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.3762,
+      "step": 2870
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5718469705290133,
+      "eval_loss": 0.4092874228954315,
+      "eval_runtime": 5.3443,
+      "eval_samples_per_second": 1022.211,
+      "eval_steps_per_second": 4.117,
+      "step": 2870
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2e-05,
+      "loss": 0.3717,
+      "step": 3280
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5736774665934469,
+      "eval_loss": 0.40997278690338135,
+      "eval_runtime": 5.339,
+      "eval_samples_per_second": 1023.223,
+      "eval_steps_per_second": 4.121,
+      "step": 3280
+    },
+    {
+      "epoch": 8.0,
+      "step": 3280,
+      "total_flos": 1.3806149824413696e+16,
+      "train_loss": 0.39033554356272626,
+      "train_runtime": 1120.6278,
+      "train_samples_per_second": 4673.407,
+      "train_steps_per_second": 18.293
+    }
+  ],
+  "max_steps": 20500,
+  "num_train_epochs": 50,
+  "total_flos": 1.3806149824413696e+16,
+  "trial_name": null,
+  "trial_params": null
+}