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 WNLI
       type: glue
       config: wnli
       split: validation
@@ -29,9 +31,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # distilbert_add_GLUE_Experiment_logit_kd_wnli_384
 
-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 WNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3444
+- Loss: 0.3434
 - Accuracy: 0.5634
 
 ## Model description

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.34341979026794434,
+    "eval_runtime": 0.0918,
+    "eval_samples": 71,
+    "eval_samples_per_second": 773.304,
+    "eval_steps_per_second": 10.892,
+    "train_loss": 0.35173843966590035,
+    "train_runtime": 21.0034,
+    "train_samples": 635,
+    "train_samples_per_second": 1511.658,
+    "train_steps_per_second": 7.142
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.34341979026794434,
+    "eval_runtime": 0.0918,
+    "eval_samples": 71,
+    "eval_samples_per_second": 773.304,
+    "eval_steps_per_second": 10.892
+}

logs/events.out.tfevents.1674957328.serv-3334.550938.23

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a65827b879d39682ef3821966c14752e069bb2ba2e6406745915f3de0baecdae
+size 357

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.0,
+    "train_loss": 0.35173843966590035,
+    "train_runtime": 21.0034,
+    "train_samples": 635,
+    "train_samples_per_second": 1511.658,
+    "train_steps_per_second": 7.142
+}

trainer_state.json

@@ -0,0 +1,160 @@
+{
+  "best_metric": 0.34341979026794434,
+  "best_model_checkpoint": "distilbert_add_GLUE_Experiment_logit_kd_wnli_384/checkpoint-12",
+  "epoch": 9.0,
+  "global_step": 27,
+  "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": 0.368,
+      "step": 3
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.34810706973075867,
+      "eval_runtime": 0.0811,
+      "eval_samples_per_second": 875.836,
+      "eval_steps_per_second": 12.336,
+      "step": 3
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3551,
+      "step": 6
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.3499087989330292,
+      "eval_runtime": 0.0812,
+      "eval_samples_per_second": 874.907,
+      "eval_steps_per_second": 12.323,
+      "step": 6
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3472,
+      "step": 9
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.3441191613674164,
+      "eval_runtime": 0.081,
+      "eval_samples_per_second": 876.571,
+      "eval_steps_per_second": 12.346,
+      "step": 9
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.3518,
+      "step": 12
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34341979026794434,
+      "eval_runtime": 0.0813,
+      "eval_samples_per_second": 873.198,
+      "eval_steps_per_second": 12.299,
+      "step": 12
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.3492,
+      "step": 15
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.3494105637073517,
+      "eval_runtime": 0.0808,
+      "eval_samples_per_second": 879.216,
+      "eval_steps_per_second": 12.383,
+      "step": 15
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.3495,
+      "step": 18
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.34805938601493835,
+      "eval_runtime": 0.0809,
+      "eval_samples_per_second": 877.583,
+      "eval_steps_per_second": 12.36,
+      "step": 18
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.3495,
+      "step": 21
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.3439640700817108,
+      "eval_runtime": 0.0826,
+      "eval_samples_per_second": 859.873,
+      "eval_steps_per_second": 12.111,
+      "step": 21
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2e-05,
+      "loss": 0.3463,
+      "step": 24
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34368786215782166,
+      "eval_runtime": 0.0814,
+      "eval_samples_per_second": 872.625,
+      "eval_steps_per_second": 12.29,
+      "step": 24
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.1e-05,
+      "loss": 0.349,
+      "step": 27
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34437814354896545,
+      "eval_runtime": 0.0806,
+      "eval_samples_per_second": 880.755,
+      "eval_steps_per_second": 12.405,
+      "step": 27
+    },
+    {
+      "epoch": 9.0,
+      "step": 27,
+      "total_flos": 149440281182208.0,
+      "train_loss": 0.35173843966590035,
+      "train_runtime": 21.0034,
+      "train_samples_per_second": 1511.658,
+      "train_steps_per_second": 7.142
+    }
+  ],
+  "max_steps": 150,
+  "num_train_epochs": 50,
+  "total_flos": 149440281182208.0,
+  "trial_name": null,
+  "trial_params": null
+}