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 MNLI
       type: glue
       config: mnli
       split: validation_matched
@@ -21,7 +23,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.817524197656648
+      value: 0.8105166802278275
 ---
 
 <!-- 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_sa_GLUE_Experiment_logit_kd_pretrain_mnli
 
-This model is a fine-tuned version of [gokuls/distilbert_sa_pre-training-complete](https://huggingface.co/gokuls/distilbert_sa_pre-training-complete) on the glue dataset.
+This model is a fine-tuned version of [gokuls/distilbert_sa_pre-training-complete](https://huggingface.co/gokuls/distilbert_sa_pre-training-complete) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3990
-- Accuracy: 0.8175
+- Loss: 0.3863
+- Accuracy: 0.8105
 
 ## Model description
 

all_results.json

@@ -0,0 +1,21 @@
+{
+    "epoch": 8.0,
+    "epoch_mm": 8.0,
+    "eval_accuracy": 0.810188487009679,
+    "eval_accuracy_mm": 0.8105166802278275,
+    "eval_loss": 0.3934803307056427,
+    "eval_loss_mm": 0.38631078600883484,
+    "eval_runtime": 11.7026,
+    "eval_runtime_mm": 11.6706,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 838.705,
+    "eval_samples_per_second_mm": 842.458,
+    "eval_steps_per_second": 3.333,
+    "eval_steps_per_second_mm": 3.342,
+    "train_loss": 0.34855276353977804,
+    "train_runtime": 6074.0966,
+    "train_samples": 392702,
+    "train_samples_per_second": 3232.596,
+    "train_steps_per_second": 12.627
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 8.0,
+    "epoch_mm": 8.0,
+    "eval_accuracy": 0.810188487009679,
+    "eval_accuracy_mm": 0.8105166802278275,
+    "eval_loss": 0.3934803307056427,
+    "eval_loss_mm": 0.38631078600883484,
+    "eval_runtime": 11.7026,
+    "eval_runtime_mm": 11.6706,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 838.705,
+    "eval_samples_per_second_mm": 842.458,
+    "eval_steps_per_second": 3.333,
+    "eval_steps_per_second_mm": 3.342
+}

logs/events.out.tfevents.1675047408.serv-3334.1119670.26

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eec33d32bd87659ed37a9c42377618be8eb26cbd170093a43d22e45ff9a1fa3f
+size 686

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 8.0,
+    "train_loss": 0.34855276353977804,
+    "train_runtime": 6074.0966,
+    "train_samples": 392702,
+    "train_samples_per_second": 3232.596,
+    "train_steps_per_second": 12.627
+}

trainer_state.json

@@ -0,0 +1,145 @@
+{
+  "best_metric": 0.3934803307056427,
+  "best_model_checkpoint": "distilbert_sa_GLUE_Experiment_logit_kd_pretrain_mnli/checkpoint-4602",
+  "epoch": 8.0,
+  "global_step": 12272,
+  "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": 0.4379,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.797554763117677,
+      "eval_loss": 0.3984040319919586,
+      "eval_runtime": 11.7481,
+      "eval_samples_per_second": 835.454,
+      "eval_steps_per_second": 3.32,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3845,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8046867040244524,
+      "eval_loss": 0.3952811658382416,
+      "eval_runtime": 11.8185,
+      "eval_samples_per_second": 830.477,
+      "eval_steps_per_second": 3.3,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.359,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.810188487009679,
+      "eval_loss": 0.3934803307056427,
+      "eval_runtime": 11.6895,
+      "eval_samples_per_second": 839.644,
+      "eval_steps_per_second": 3.336,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.3411,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.807743250127356,
+      "eval_loss": 0.3962075114250183,
+      "eval_runtime": 11.855,
+      "eval_samples_per_second": 827.919,
+      "eval_steps_per_second": 3.29,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.50006518904824e-05,
+      "loss": 0.3279,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8172185430463577,
+      "eval_loss": 0.3958968222141266,
+      "eval_runtime": 11.6839,
+      "eval_samples_per_second": 840.044,
+      "eval_steps_per_second": 3.338,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.40006518904824e-05,
+      "loss": 0.3189,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.810188487009679,
+      "eval_loss": 0.40176525712013245,
+      "eval_runtime": 11.6931,
+      "eval_samples_per_second": 839.381,
+      "eval_steps_per_second": 3.335,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3001303780964804e-05,
+      "loss": 0.3119,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.807335710646969,
+      "eval_loss": 0.4040135145187378,
+      "eval_runtime": 11.7329,
+      "eval_samples_per_second": 836.534,
+      "eval_steps_per_second": 3.324,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.20019556714472e-05,
+      "loss": 0.3071,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.817524197656648,
+      "eval_loss": 0.3989511728286743,
+      "eval_runtime": 11.7287,
+      "eval_samples_per_second": 836.837,
+      "eval_steps_per_second": 3.325,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "step": 12272,
+      "total_flos": 2.0808456842143334e+17,
+      "train_loss": 0.34855276353977804,
+      "train_runtime": 6074.0966,
+      "train_samples_per_second": 3232.596,
+      "train_steps_per_second": 12.627
+    }
+  ],
+  "max_steps": 76700,
+  "num_train_epochs": 50,
+  "total_flos": 2.0808456842143334e+17,
+  "trial_name": null,
+  "trial_params": null
+}