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. -->
 
 # mobilebert_add_GLUE_Experiment_logit_kd_wnli
 
-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 WNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3457
+- Loss: 0.3448
 - Accuracy: 0.5634
 
 ## Model description

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.34482690691947937,
+    "eval_runtime": 0.217,
+    "eval_samples": 71,
+    "eval_samples_per_second": 327.209,
+    "eval_steps_per_second": 4.609,
+    "train_loss": 0.3469098753399319,
+    "train_runtime": 81.7482,
+    "train_samples": 635,
+    "train_samples_per_second": 388.388,
+    "train_steps_per_second": 3.058
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.5633802816901409,
+    "eval_loss": 0.34482690691947937,
+    "eval_runtime": 0.217,
+    "eval_samples": 71,
+    "eval_samples_per_second": 327.209,
+    "eval_steps_per_second": 4.609
+}

logs/events.out.tfevents.1674985793.serv-3332.3626077.23

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d39bb5a20bca4ae7f25487af5cb53ed06407eeb9bd4393f7efc21aeb843d95ca
+size 357

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.0,
+    "train_loss": 0.3469098753399319,
+    "train_runtime": 81.7482,
+    "train_samples": 635,
+    "train_samples_per_second": 388.388,
+    "train_steps_per_second": 3.058
+}

trainer_state.json

@@ -0,0 +1,160 @@
+{
+  "best_metric": 0.34482690691947937,
+  "best_model_checkpoint": "mobilebert_add_GLUE_Experiment_logit_kd_wnli/checkpoint-20",
+  "epoch": 9.0,
+  "global_step": 45,
+  "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": 0.3478,
+      "step": 5
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34598228335380554,
+      "eval_runtime": 0.2035,
+      "eval_samples_per_second": 348.952,
+      "eval_steps_per_second": 4.915,
+      "step": 5
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3477,
+      "step": 10
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.43661971830985913,
+      "eval_loss": 0.34800270199775696,
+      "eval_runtime": 0.2032,
+      "eval_samples_per_second": 349.476,
+      "eval_steps_per_second": 4.922,
+      "step": 10
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3466,
+      "step": 15
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34593814611434937,
+      "eval_runtime": 0.2041,
+      "eval_samples_per_second": 347.936,
+      "eval_steps_per_second": 4.901,
+      "step": 15
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.3466,
+      "step": 20
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34482690691947937,
+      "eval_runtime": 0.2015,
+      "eval_samples_per_second": 352.284,
+      "eval_steps_per_second": 4.962,
+      "step": 20
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.3468,
+      "step": 25
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.3451252579689026,
+      "eval_runtime": 0.2027,
+      "eval_samples_per_second": 350.339,
+      "eval_steps_per_second": 4.934,
+      "step": 25
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.3467,
+      "step": 30
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.3461253345012665,
+      "eval_runtime": 0.2015,
+      "eval_samples_per_second": 352.419,
+      "eval_steps_per_second": 4.964,
+      "step": 30
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.3465,
+      "step": 35
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34646010398864746,
+      "eval_runtime": 0.2846,
+      "eval_samples_per_second": 249.512,
+      "eval_steps_per_second": 3.514,
+      "step": 35
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2e-05,
+      "loss": 0.3466,
+      "step": 40
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.34657594561576843,
+      "eval_runtime": 0.202,
+      "eval_samples_per_second": 351.433,
+      "eval_steps_per_second": 4.95,
+      "step": 40
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.1e-05,
+      "loss": 0.3468,
+      "step": 45
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.5633802816901409,
+      "eval_loss": 0.345743864774704,
+      "eval_runtime": 0.2015,
+      "eval_samples_per_second": 352.344,
+      "eval_steps_per_second": 4.963,
+      "step": 45
+    },
+    {
+      "epoch": 9.0,
+      "step": 45,
+      "total_flos": 169051787624448.0,
+      "train_loss": 0.3469098753399319,
+      "train_runtime": 81.7482,
+      "train_samples_per_second": 388.388,
+      "train_steps_per_second": 3.058
+    }
+  ],
+  "max_steps": 250,
+  "num_train_epochs": 50,
+  "total_flos": 169051787624448.0,
+  "trial_name": null,
+  "trial_params": null
+}