End of training

README.md

@@ -1,14 +1,32 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/tiny_bert_km_10_v2
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 - f1
 model-index:
 - name: tiny_bert_km_10_v2_qqp
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE QQP
+      type: glue
+      args: qqp
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.8163987138263665
+    - name: F1
+      type: f1
+      value: 0.7454302273740526
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -16,12 +34,12 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tiny_bert_km_10_v2_qqp
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_km_10_v2](https://huggingface.co/Hartunka/tiny_bert_km_10_v2) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_km_10_v2](https://huggingface.co/Hartunka/tiny_bert_km_10_v2) on the GLUE QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5614
-- Accuracy: 0.8253
-- F1: 0.7663
-- Combined Score: 0.7958
+- Loss: 0.4118
+- Accuracy: 0.8164
+- F1: 0.7454
+- Combined Score: 0.7809
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8163987138263665,
+    "eval_combined_score": 0.7809144706002096,
+    "eval_f1": 0.7454302273740526,
+    "eval_loss": 0.4117882549762726,
+    "eval_runtime": 11.5329,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 3505.635,
+    "eval_steps_per_second": 13.7,
+    "total_flos": 7.633075201391002e+16,
+    "train_loss": 0.2866973742821716,
+    "train_runtime": 1610.9558,
+    "train_samples": 363846,
+    "train_samples_per_second": 11292.861,
+    "train_steps_per_second": 44.135
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8163987138263665,
+    "eval_combined_score": 0.7809144706002096,
+    "eval_f1": 0.7454302273740526,
+    "eval_loss": 0.4117882549762726,
+    "eval_runtime": 11.5329,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 3505.635,
+    "eval_steps_per_second": 13.7
+}

logs/events.out.tfevents.1745275725.s_005_m.2836625.25

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:075add1d2d5dce83bc64c2cd4a1bdc9a91708e9bbae26586b85ca9fd2fa1a9fc
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "total_flos": 7.633075201391002e+16,
+    "train_loss": 0.2866973742821716,
+    "train_runtime": 1610.9558,
+    "train_samples": 363846,
+    "train_samples_per_second": 11292.861,
+    "train_steps_per_second": 44.135
+}

trainer_state.json

@@ -0,0 +1,196 @@
+{
+  "best_global_step": 4266,
+  "best_metric": 0.4117882549762726,
+  "best_model_checkpoint": "tiny_bert_km_10_v2_qqp/checkpoint-4266",
+  "epoch": 8.0,
+  "eval_steps": 500,
+  "global_step": 11376,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5244591236114502,
+      "learning_rate": 4.9e-05,
+      "loss": 0.49,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7852337373237694,
+      "eval_combined_score": 0.7324434689201397,
+      "eval_f1": 0.6796532005165099,
+      "eval_loss": 0.4448812007904053,
+      "eval_runtime": 11.6504,
+      "eval_samples_per_second": 3470.282,
+      "eval_steps_per_second": 13.562,
+      "step": 1422
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.719128131866455,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3992,
+      "step": 2844
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8037843185753154,
+      "eval_combined_score": 0.7603881101991485,
+      "eval_f1": 0.7169919018229817,
+      "eval_loss": 0.42040717601776123,
+      "eval_runtime": 11.4731,
+      "eval_samples_per_second": 3523.884,
+      "eval_steps_per_second": 13.771,
+      "step": 2844
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.167677640914917,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3367,
+      "step": 4266
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8163987138263665,
+      "eval_combined_score": 0.7809144706002096,
+      "eval_f1": 0.7454302273740526,
+      "eval_loss": 0.4117882549762726,
+      "eval_runtime": 11.456,
+      "eval_samples_per_second": 3529.152,
+      "eval_steps_per_second": 13.792,
+      "step": 4266
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.39666748046875,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.2845,
+      "step": 5688
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8199604254266634,
+      "eval_combined_score": 0.7824440316151083,
+      "eval_f1": 0.7449276378035533,
+      "eval_loss": 0.42973750829696655,
+      "eval_runtime": 11.7458,
+      "eval_samples_per_second": 3442.082,
+      "eval_steps_per_second": 13.452,
+      "step": 5688
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 3.2635037899017334,
+      "learning_rate": 4.5e-05,
+      "loss": 0.243,
+      "step": 7110
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8228295819935691,
+      "eval_combined_score": 0.7932485697423717,
+      "eval_f1": 0.7636675574911742,
+      "eval_loss": 0.4527283310890198,
+      "eval_runtime": 11.4678,
+      "eval_samples_per_second": 3525.524,
+      "eval_steps_per_second": 13.778,
+      "step": 7110
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 4.172083377838135,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.2066,
+      "step": 8532
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8198862231016572,
+      "eval_combined_score": 0.7917595228725631,
+      "eval_f1": 0.7636328226434692,
+      "eval_loss": 0.4895334541797638,
+      "eval_runtime": 11.4302,
+      "eval_samples_per_second": 3537.129,
+      "eval_steps_per_second": 13.823,
+      "step": 8532
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 5.709171295166016,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1782,
+      "step": 9954
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8204551075933713,
+      "eval_combined_score": 0.7936545651986449,
+      "eval_f1": 0.7668540228039185,
+      "eval_loss": 0.5355231761932373,
+      "eval_runtime": 11.456,
+      "eval_samples_per_second": 3529.158,
+      "eval_steps_per_second": 13.792,
+      "step": 9954
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.7236862182617188,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1554,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8252782587187731,
+      "eval_combined_score": 0.7957783564636746,
+      "eval_f1": 0.766278454208576,
+      "eval_loss": 0.5613935589790344,
+      "eval_runtime": 11.5118,
+      "eval_samples_per_second": 3512.059,
+      "eval_steps_per_second": 13.725,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "step": 11376,
+      "total_flos": 7.633075201391002e+16,
+      "train_loss": 0.2866973742821716,
+      "train_runtime": 1610.9558,
+      "train_samples_per_second": 11292.861,
+      "train_steps_per_second": 44.135
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 71100,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 5
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.633075201391002e+16,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}