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: 0.7908978481325748
+      value: 0.7871877318822657
     - name: F1
       type: f1
-      value: 0.721137353212825
+      value: 0.7061676115019466
 ---
 
 <!-- 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_128
 
-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: 0.7410
-- Accuracy: 0.7909
-- F1: 0.7211
-- Combined Score: 0.7560
+- Loss: 0.6884
+- Accuracy: 0.7872
+- F1: 0.7062
+- Combined Score: 0.7467
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 16.0,
+    "eval_accuracy": 0.7871877318822657,
+    "eval_combined_score": 0.7466776716921062,
+    "eval_f1": 0.7061676115019466,
+    "eval_loss": 0.6884405016899109,
+    "eval_runtime": 71.7576,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 563.425,
+    "eval_steps_per_second": 4.404,
+    "train_loss": 0.6153143427550604,
+    "train_runtime": 24971.032,
+    "train_samples": 363846,
+    "train_samples_per_second": 728.536,
+    "train_steps_per_second": 5.693
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 16.0,
+    "eval_accuracy": 0.7871877318822657,
+    "eval_combined_score": 0.7466776716921062,
+    "eval_f1": 0.7061676115019466,
+    "eval_loss": 0.6884405016899109,
+    "eval_runtime": 71.7576,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 563.425,
+    "eval_steps_per_second": 4.404
+}

logs/events.out.tfevents.1675058847.serv-3334.1021177.11

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba7e5388504a46a6906183e587223e79a40384f2576deb3a1f048a13262ec713
+size 475

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 16.0,
+    "train_loss": 0.6153143427550604,
+    "train_runtime": 24971.032,
+    "train_samples": 363846,
+    "train_samples_per_second": 728.536,
+    "train_steps_per_second": 5.693
+}

trainer_state.json

@@ -0,0 +1,297 @@
+{
+  "best_metric": 0.6884405016899109,
+  "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_qqp_128/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": 0.9518,
+      "step": 2843
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7536482809794707,
+      "eval_combined_score": 0.7033049767266691,
+      "eval_f1": 0.6529616724738676,
+      "eval_loss": 0.8352172374725342,
+      "eval_runtime": 71.9208,
+      "eval_samples_per_second": 562.146,
+      "eval_steps_per_second": 4.394,
+      "step": 2843
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 0.8249,
+      "step": 5686
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7606727677467228,
+      "eval_combined_score": 0.6912633809032191,
+      "eval_f1": 0.6218539940597155,
+      "eval_loss": 0.7766293883323669,
+      "eval_runtime": 71.8035,
+      "eval_samples_per_second": 563.064,
+      "eval_steps_per_second": 4.401,
+      "step": 5686
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 0.7847,
+      "step": 8529
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.764828097947069,
+      "eval_combined_score": 0.7025018479441673,
+      "eval_f1": 0.6401755979412656,
+      "eval_loss": 0.7624924778938293,
+      "eval_runtime": 71.7415,
+      "eval_samples_per_second": 563.551,
+      "eval_steps_per_second": 4.405,
+      "step": 8529
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.7498,
+      "step": 11372
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7637892653969824,
+      "eval_combined_score": 0.6917464963493354,
+      "eval_f1": 0.6197037273016884,
+      "eval_loss": 0.7550554275512695,
+      "eval_runtime": 71.5767,
+      "eval_samples_per_second": 564.848,
+      "eval_steps_per_second": 4.415,
+      "step": 11372
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 0.7137,
+      "step": 14215
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.7691318327974277,
+      "eval_combined_score": 0.7118036449838919,
+      "eval_f1": 0.6544754571703562,
+      "eval_loss": 0.7386809587478638,
+      "eval_runtime": 71.6477,
+      "eval_samples_per_second": 564.289,
+      "eval_steps_per_second": 4.41,
+      "step": 14215
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.6762,
+      "step": 17058
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.7753153598812763,
+      "eval_combined_score": 0.7236628792473973,
+      "eval_f1": 0.6720103986135182,
+      "eval_loss": 0.7165194153785706,
+      "eval_runtime": 71.8883,
+      "eval_samples_per_second": 562.4,
+      "eval_steps_per_second": 4.396,
+      "step": 17058
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6373,
+      "step": 19901
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.778308186989859,
+      "eval_combined_score": 0.7273846100454836,
+      "eval_f1": 0.6764610331011082,
+      "eval_loss": 0.7042196989059448,
+      "eval_runtime": 71.6003,
+      "eval_samples_per_second": 564.662,
+      "eval_steps_per_second": 4.413,
+      "step": 19901
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6045,
+      "step": 22744
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7798664358149888,
+      "eval_combined_score": 0.7350428593972265,
+      "eval_f1": 0.690219282979464,
+      "eval_loss": 0.7075350880622864,
+      "eval_runtime": 71.5758,
+      "eval_samples_per_second": 564.856,
+      "eval_steps_per_second": 4.415,
+      "step": 22744
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.1e-05,
+      "loss": 0.5729,
+      "step": 25587
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.7791738807815978,
+      "eval_combined_score": 0.7215273063270988,
+      "eval_f1": 0.6638807318725999,
+      "eval_loss": 0.7232872843742371,
+      "eval_runtime": 71.3607,
+      "eval_samples_per_second": 566.558,
+      "eval_steps_per_second": 4.428,
+      "step": 25587
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4e-05,
+      "loss": 0.545,
+      "step": 28430
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.7805342567400445,
+      "eval_combined_score": 0.7492781557955792,
+      "eval_f1": 0.7180220548511138,
+      "eval_loss": 0.708789587020874,
+      "eval_runtime": 71.5867,
+      "eval_samples_per_second": 564.77,
+      "eval_steps_per_second": 4.414,
+      "step": 28430
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.5183,
+      "step": 31273
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.7871877318822657,
+      "eval_combined_score": 0.7466776716921062,
+      "eval_f1": 0.7061676115019466,
+      "eval_loss": 0.6884405016899109,
+      "eval_runtime": 71.6715,
+      "eval_samples_per_second": 564.102,
+      "eval_steps_per_second": 4.409,
+      "step": 31273
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 3.8e-05,
+      "loss": 0.4948,
+      "step": 34116
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.7868661884739055,
+      "eval_combined_score": 0.7472385300891464,
+      "eval_f1": 0.7076108717043874,
+      "eval_loss": 0.7063745260238647,
+      "eval_runtime": 71.4754,
+      "eval_samples_per_second": 565.649,
+      "eval_steps_per_second": 4.421,
+      "step": 34116
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 3.7e-05,
+      "loss": 0.4724,
+      "step": 36959
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.7883502349740292,
+      "eval_combined_score": 0.7501615189141915,
+      "eval_f1": 0.7119728028543537,
+      "eval_loss": 0.7052645683288574,
+      "eval_runtime": 71.6405,
+      "eval_samples_per_second": 564.345,
+      "eval_steps_per_second": 4.411,
+      "step": 36959
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.4514,
+      "step": 39802
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.7903042295325253,
+      "eval_combined_score": 0.7463735102456954,
+      "eval_f1": 0.7024427909588656,
+      "eval_loss": 0.7314090728759766,
+      "eval_runtime": 71.7172,
+      "eval_samples_per_second": 563.742,
+      "eval_steps_per_second": 4.406,
+      "step": 39802
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.5e-05,
+      "loss": 0.4321,
+      "step": 42645
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.7891169923324264,
+      "eval_combined_score": 0.7559694194431963,
+      "eval_f1": 0.7228218465539662,
+      "eval_loss": 0.7111542820930481,
+      "eval_runtime": 71.6308,
+      "eval_samples_per_second": 564.422,
+      "eval_steps_per_second": 4.412,
+      "step": 42645
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.4152,
+      "step": 45488
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.7908978481325748,
+      "eval_combined_score": 0.7560176006727,
+      "eval_f1": 0.721137353212825,
+      "eval_loss": 0.7409694194793701,
+      "eval_runtime": 71.6151,
+      "eval_samples_per_second": 564.546,
+      "eval_steps_per_second": 4.412,
+      "step": 45488
+    },
+    {
+      "epoch": 16.0,
+      "step": 45488,
+      "total_flos": 1.3875395219895091e+17,
+      "train_loss": 0.6153143427550604,
+      "train_runtime": 24971.032,
+      "train_samples_per_second": 728.536,
+      "train_steps_per_second": 5.693
+    }
+  ],
+  "max_steps": 142150,
+  "num_train_epochs": 50,
+  "total_flos": 1.3875395219895091e+17,
+  "trial_name": null,
+  "trial_params": null
+}