End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 license: apache-2.0
 tags:
 - generated_from_trainer
@@ -14,7 +16,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE QQP
       type: glue
       config: qqp
       split: validation
@@ -22,10 +24,10 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.7957209992579768
+      value: 0.794855305466238
     - name: F1
       type: f1
-      value: 0.729983326249714
+      value: 0.7224044447419508
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -33,12 +35,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 # mobilebert_sa_GLUE_Experiment_logit_kd_qqp_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 QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7136
-- Accuracy: 0.7957
-- F1: 0.7300
-- Combined Score: 0.7629
+- Loss: 0.6619
+- Accuracy: 0.7949
+- F1: 0.7224
+- Combined Score: 0.7586
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 16.0,
+    "eval_accuracy": 0.794855305466238,
+    "eval_combined_score": 0.7586298751040943,
+    "eval_f1": 0.7224044447419508,
+    "eval_loss": 0.6618520021438599,
+    "eval_runtime": 72.2937,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 559.246,
+    "eval_steps_per_second": 4.371,
+    "train_loss": 0.5890673631687546,
+    "train_runtime": 25425.7527,
+    "train_samples": 363846,
+    "train_samples_per_second": 715.507,
+    "train_steps_per_second": 5.591
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 16.0,
+    "eval_accuracy": 0.794855305466238,
+    "eval_combined_score": 0.7586298751040943,
+    "eval_f1": 0.7224044447419508,
+    "eval_loss": 0.6618520021438599,
+    "eval_runtime": 72.2937,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 559.246,
+    "eval_steps_per_second": 4.371
+}

logs/events.out.tfevents.1675059329.serv-3334.1016902.11

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6ec17e84c161c68c06ca1514d71d8f4121d67fa9e5b4fc8f8c9d092142530f7
+size 475

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 16.0,
+    "train_loss": 0.5890673631687546,
+    "train_runtime": 25425.7527,
+    "train_samples": 363846,
+    "train_samples_per_second": 715.507,
+    "train_steps_per_second": 5.591
+}

trainer_state.json

@@ -0,0 +1,297 @@
+{
+  "best_metric": 0.6618520021438599,
+  "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_qqp_256/checkpoint-31273",
+  "epoch": 16.0,
+  "global_step": 45488,
+  "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": 0.9454,
+      "step": 2843
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.755552807321296,
+      "eval_combined_score": 0.7059041209574377,
+      "eval_f1": 0.6562554345935794,
+      "eval_loss": 0.8257145881652832,
+      "eval_runtime": 72.5003,
+      "eval_samples_per_second": 557.653,
+      "eval_steps_per_second": 4.359,
+      "step": 2843
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.8165,
+      "step": 5686
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7589661142715806,
+      "eval_combined_score": 0.6819519426913909,
+      "eval_f1": 0.6049377711112012,
+      "eval_loss": 0.7681939005851746,
+      "eval_runtime": 72.4829,
+      "eval_samples_per_second": 557.787,
+      "eval_steps_per_second": 4.36,
+      "step": 5686
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.7741,
+      "step": 8529
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7638139995053178,
+      "eval_combined_score": 0.6920433977325372,
+      "eval_f1": 0.6202727959597566,
+      "eval_loss": 0.7514052987098694,
+      "eval_runtime": 72.7836,
+      "eval_samples_per_second": 555.482,
+      "eval_steps_per_second": 4.342,
+      "step": 8529
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.7325,
+      "step": 11372
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7674746475389562,
+      "eval_combined_score": 0.698115556996189,
+      "eval_f1": 0.6287564664534218,
+      "eval_loss": 0.7353646755218506,
+      "eval_runtime": 72.4082,
+      "eval_samples_per_second": 558.362,
+      "eval_steps_per_second": 4.364,
+      "step": 11372
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6849,
+      "step": 14215
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.7785060598565422,
+      "eval_combined_score": 0.7301661583702679,
+      "eval_f1": 0.6818262568839936,
+      "eval_loss": 0.7063168883323669,
+      "eval_runtime": 72.761,
+      "eval_samples_per_second": 555.655,
+      "eval_steps_per_second": 4.343,
+      "step": 14215
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.6399,
+      "step": 17058
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.78276032649023,
+      "eval_combined_score": 0.7351824040418427,
+      "eval_f1": 0.6876044815934554,
+      "eval_loss": 0.6906189918518066,
+      "eval_runtime": 72.4222,
+      "eval_samples_per_second": 558.254,
+      "eval_steps_per_second": 4.363,
+      "step": 17058
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6005,
+      "step": 19901
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7867919861488993,
+      "eval_combined_score": 0.7430436592992455,
+      "eval_f1": 0.6992953324495919,
+      "eval_loss": 0.677066445350647,
+      "eval_runtime": 72.5064,
+      "eval_samples_per_second": 557.606,
+      "eval_steps_per_second": 4.358,
+      "step": 19901
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2e-05,
+      "loss": 0.5666,
+      "step": 22744
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7897106109324759,
+      "eval_combined_score": 0.7517336301581778,
+      "eval_f1": 0.71375664938388,
+      "eval_loss": 0.6808561086654663,
+      "eval_runtime": 72.3704,
+      "eval_samples_per_second": 558.654,
+      "eval_steps_per_second": 4.366,
+      "step": 22744
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.1e-05,
+      "loss": 0.5365,
+      "step": 25587
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.7885975760573831,
+      "eval_combined_score": 0.7403265307777458,
+      "eval_f1": 0.6920554854981085,
+      "eval_loss": 0.6807268857955933,
+      "eval_runtime": 72.5072,
+      "eval_samples_per_second": 557.6,
+      "eval_steps_per_second": 4.358,
+      "step": 25587
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4e-05,
+      "loss": 0.5097,
+      "step": 28430
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.7872619342072719,
+      "eval_combined_score": 0.7566418293977779,
+      "eval_f1": 0.726021724588284,
+      "eval_loss": 0.6826764345169067,
+      "eval_runtime": 72.5186,
+      "eval_samples_per_second": 557.512,
+      "eval_steps_per_second": 4.358,
+      "step": 28430
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.4856,
+      "step": 31273
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.794855305466238,
+      "eval_combined_score": 0.7586298751040943,
+      "eval_f1": 0.7224044447419508,
+      "eval_loss": 0.6618520021438599,
+      "eval_runtime": 72.4772,
+      "eval_samples_per_second": 557.83,
+      "eval_steps_per_second": 4.36,
+      "step": 31273
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 3.8e-05,
+      "loss": 0.4653,
+      "step": 34116
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.7947563690328964,
+      "eval_combined_score": 0.7572282047839796,
+      "eval_f1": 0.7197000405350629,
+      "eval_loss": 0.7001789212226868,
+      "eval_runtime": 72.5909,
+      "eval_samples_per_second": 556.957,
+      "eval_steps_per_second": 4.353,
+      "step": 34116
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 3.7e-05,
+      "loss": 0.4438,
+      "step": 36959
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.7965372248330448,
+      "eval_combined_score": 0.7584372396008563,
+      "eval_f1": 0.7203372543686679,
+      "eval_loss": 0.689961850643158,
+      "eval_runtime": 72.5363,
+      "eval_samples_per_second": 557.376,
+      "eval_steps_per_second": 4.356,
+      "step": 36959
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.4255,
+      "step": 39802
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.7980707395498392,
+      "eval_combined_score": 0.7632222209859922,
+      "eval_f1": 0.7283737024221453,
+      "eval_loss": 0.6847425699234009,
+      "eval_runtime": 72.6156,
+      "eval_samples_per_second": 556.767,
+      "eval_steps_per_second": 4.352,
+      "step": 39802
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.5e-05,
+      "loss": 0.4072,
+      "step": 42645
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.7917388078159783,
+      "eval_combined_score": 0.7626668474313707,
+      "eval_f1": 0.7335948870467632,
+      "eval_loss": 0.6892824769020081,
+      "eval_runtime": 72.7761,
+      "eval_samples_per_second": 555.539,
+      "eval_steps_per_second": 4.342,
+      "step": 42645
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.391,
+      "step": 45488
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.7957209992579768,
+      "eval_combined_score": 0.7628521627538454,
+      "eval_f1": 0.729983326249714,
+      "eval_loss": 0.7136414647102356,
+      "eval_runtime": 72.54,
+      "eval_samples_per_second": 557.348,
+      "eval_steps_per_second": 4.356,
+      "step": 45488
+    },
+    {
+      "epoch": 16.0,
+      "step": 45488,
+      "total_flos": 1.5334594896828826e+17,
+      "train_loss": 0.5890673631687546,
+      "train_runtime": 25425.7527,
+      "train_samples_per_second": 715.507,
+      "train_steps_per_second": 5.591
+    }
+  ],
+  "max_steps": 142150,
+  "num_train_epochs": 50,
+  "total_flos": 1.5334594896828826e+17,
+  "trial_name": null,
+  "trial_params": null
+}