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: 0.5667215815485996
+      value: 0.5870400878638111
 ---
 
 <!-- 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_qnli_192
 
-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: 0.4269
-- Accuracy: 0.5667
+- Loss: 0.3931
+- Accuracy: 0.5870
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.5870400878638111,
+    "eval_loss": 0.3931228816509247,
+    "eval_runtime": 7.3569,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 742.565,
+    "eval_steps_per_second": 2.99,
+    "train_loss": 0.37907483569421,
+    "train_runtime": 1348.5756,
+    "train_samples": 104743,
+    "train_samples_per_second": 3883.468,
+    "train_steps_per_second": 15.201
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.5870400878638111,
+    "eval_loss": 0.3931228816509247,
+    "eval_runtime": 7.3569,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 742.565,
+    "eval_steps_per_second": 2.99
+}

logs/events.out.tfevents.1675031721.serv-3332.597879.8

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:570222e3b070bd2802b815ce8df003a5ce3136c98d3edba783230681445047d7
+size 363

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 7.0,
+    "train_loss": 0.37907483569421,
+    "train_runtime": 1348.5756,
+    "train_samples": 104743,
+    "train_samples_per_second": 3883.468,
+    "train_steps_per_second": 15.201
+}

trainer_state.json

@@ -0,0 +1,130 @@
+{
+  "best_metric": 0.3931228816509247,
+  "best_model_checkpoint": "distilbert_sa_GLUE_Experiment_logit_kd_qnli_192/checkpoint-820",
+  "epoch": 7.0,
+  "global_step": 2870,
+  "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": 0.4083,
+      "step": 410
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5734944169870034,
+      "eval_loss": 0.3945602476596832,
+      "eval_runtime": 7.2349,
+      "eval_samples_per_second": 755.094,
+      "eval_steps_per_second": 3.041,
+      "step": 410
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3936,
+      "step": 820
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5870400878638111,
+      "eval_loss": 0.3931228816509247,
+      "eval_runtime": 7.4099,
+      "eval_samples_per_second": 737.261,
+      "eval_steps_per_second": 2.969,
+      "step": 820
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3843,
+      "step": 1230
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5863078894380377,
+      "eval_loss": 0.3935385048389435,
+      "eval_runtime": 7.2662,
+      "eval_samples_per_second": 751.832,
+      "eval_steps_per_second": 3.028,
+      "step": 1230
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.3766,
+      "step": 1640
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5857587406187077,
+      "eval_loss": 0.39795002341270447,
+      "eval_runtime": 7.2532,
+      "eval_samples_per_second": 753.18,
+      "eval_steps_per_second": 3.033,
+      "step": 1640
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.3699,
+      "step": 2050
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5780706571480871,
+      "eval_loss": 0.39961472153663635,
+      "eval_runtime": 7.2932,
+      "eval_samples_per_second": 749.058,
+      "eval_steps_per_second": 3.017,
+      "step": 2050
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.3636,
+      "step": 2460
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5795350539996339,
+      "eval_loss": 0.41122666001319885,
+      "eval_runtime": 7.2446,
+      "eval_samples_per_second": 754.08,
+      "eval_steps_per_second": 3.037,
+      "step": 2460
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.3572,
+      "step": 2870
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5667215815485996,
+      "eval_loss": 0.42693567276000977,
+      "eval_runtime": 7.3154,
+      "eval_samples_per_second": 746.778,
+      "eval_steps_per_second": 3.007,
+      "step": 2870
+    },
+    {
+      "epoch": 7.0,
+      "step": 2870,
+      "total_flos": 9042515428638720.0,
+      "train_loss": 0.37907483569421,
+      "train_runtime": 1348.5756,
+      "train_samples_per_second": 3883.468,
+      "train_steps_per_second": 15.201
+    }
+  ],
+  "max_steps": 20500,
+  "num_train_epochs": 50,
+  "total_flos": 9042515428638720.0,
+  "trial_name": null,
+  "trial_params": null
+}