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: 0.5967417170053084
+      value: 0.6163280248947465
 ---
 
 <!-- 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. -->
 
 # mobilebert_sa_GLUE_Experiment_logit_kd_qnli_256
 
-This model is a fine-tuned version of [google/mobilebert-uncased](https://huggingface.co/google/mobilebert-uncased) on the glue dataset.
+This model is a fine-tuned version of [google/mobilebert-uncased](https://huggingface.co/google/mobilebert-uncased) on the GLUE QNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0317
-- Accuracy: 0.5967
+- Loss: 0.9616
+- Accuracy: 0.6163
 
 ## Model description
 

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.6163280248947465,
+    "eval_loss": 0.9615830183029175,
+    "eval_runtime": 9.7721,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 559.041,
+    "eval_steps_per_second": 4.4,
+    "train_loss": 0.9618304146328003,
+    "train_runtime": 3166.4896,
+    "train_samples": 104743,
+    "train_samples_per_second": 1653.929,
+    "train_steps_per_second": 12.932
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.6163280248947465,
+    "eval_loss": 0.9615830183029175,
+    "eval_runtime": 9.7721,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 559.041,
+    "eval_steps_per_second": 4.4
+}

logs/events.out.tfevents.1675033618.serv-3334.1016902.8

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e569f1c087deb9e08f4933804a9a7a4036a592907f594a5c6520680c48061652
+size 363

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 7.0,
+    "train_loss": 0.9618304146328003,
+    "train_runtime": 3166.4896,
+    "train_samples": 104743,
+    "train_samples_per_second": 1653.929,
+    "train_steps_per_second": 12.932
+}

trainer_state.json

@@ -0,0 +1,130 @@
+{
+  "best_metric": 0.9615830183029175,
+  "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_qnli_256/checkpoint-1638",
+  "epoch": 7.0,
+  "global_step": 5733,
+  "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": 1.1009,
+      "step": 819
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5987552626761853,
+      "eval_loss": 0.9864959716796875,
+      "eval_runtime": 9.8348,
+      "eval_samples_per_second": 555.476,
+      "eval_steps_per_second": 4.372,
+      "step": 819
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 1.019,
+      "step": 1638
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6163280248947465,
+      "eval_loss": 0.9615830183029175,
+      "eval_runtime": 9.7787,
+      "eval_samples_per_second": 558.662,
+      "eval_steps_per_second": 4.397,
+      "step": 1638
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.9743,
+      "step": 2457
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.613399231191653,
+      "eval_loss": 0.9671955704689026,
+      "eval_runtime": 9.8271,
+      "eval_samples_per_second": 555.911,
+      "eval_steps_per_second": 4.376,
+      "step": 2457
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.942,
+      "step": 3276
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6069924949661358,
+      "eval_loss": 0.9723971486091614,
+      "eval_runtime": 9.779,
+      "eval_samples_per_second": 558.647,
+      "eval_steps_per_second": 4.397,
+      "step": 3276
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.9189,
+      "step": 4095
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6016840563792788,
+      "eval_loss": 0.9827193021774292,
+      "eval_runtime": 9.8296,
+      "eval_samples_per_second": 555.769,
+      "eval_steps_per_second": 4.375,
+      "step": 4095
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.898,
+      "step": 4914
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5958264689730917,
+      "eval_loss": 1.0090252161026,
+      "eval_runtime": 9.7883,
+      "eval_samples_per_second": 558.113,
+      "eval_steps_per_second": 4.393,
+      "step": 4914
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.8798,
+      "step": 5733
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5967417170053084,
+      "eval_loss": 1.0317091941833496,
+      "eval_runtime": 9.8635,
+      "eval_samples_per_second": 553.859,
+      "eval_steps_per_second": 4.359,
+      "step": 5733
+    },
+    {
+      "epoch": 7.0,
+      "step": 5733,
+      "total_flos": 1.931335714419507e+16,
+      "train_loss": 0.9618304146328003,
+      "train_runtime": 3166.4896,
+      "train_samples_per_second": 1653.929,
+      "train_steps_per_second": 12.932
+    }
+  ],
+  "max_steps": 40950,
+  "num_train_epochs": 50,
+  "total_flos": 1.931335714419507e+16,
+  "trial_name": null,
+  "trial_params": null
+}