End of training

README.md

@@ -1,14 +1,32 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/distilbert_km_50_v2
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 - f1
 model-index:
 - name: distilbert_km_50_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: 0.8210734603017561
+    - name: F1
+      type: f1
+      value: 0.7656776366934439
 ---
 
 <!-- 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. -->
 
 # distilbert_km_50_v2_qqp
 
-This model is a fine-tuned version of [Hartunka/distilbert_km_50_v2](https://huggingface.co/Hartunka/distilbert_km_50_v2) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/distilbert_km_50_v2](https://huggingface.co/Hartunka/distilbert_km_50_v2) on the GLUE QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5893
-- Accuracy: 0.8175
-- F1: 0.7745
-- Combined Score: 0.7960
+- Loss: 0.4002
+- Accuracy: 0.8211
+- F1: 0.7657
+- Combined Score: 0.7934
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8210734603017561,
+    "eval_combined_score": 0.7933755484976,
+    "eval_f1": 0.7656776366934439,
+    "eval_loss": 0.40021559596061707,
+    "eval_runtime": 16.4127,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 2463.34,
+    "eval_steps_per_second": 9.627,
+    "total_flos": 1.927909325255639e+17,
+    "train_loss": 0.269940188497766,
+    "train_runtime": 2790.3681,
+    "train_samples": 363846,
+    "train_samples_per_second": 6519.677,
+    "train_steps_per_second": 25.481
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8210734603017561,
+    "eval_combined_score": 0.7933755484976,
+    "eval_f1": 0.7656776366934439,
+    "eval_loss": 0.40021559596061707,
+    "eval_runtime": 16.4127,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 2463.34,
+    "eval_steps_per_second": 9.627
+}

logs/events.out.tfevents.1745333937.s_005_m.2850345.61

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:108ea6b604f25df07b591aa24a88c873a71c3f3dd037fc88ace4d91c54142531
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "total_flos": 1.927909325255639e+17,
+    "train_loss": 0.269940188497766,
+    "train_runtime": 2790.3681,
+    "train_samples": 363846,
+    "train_samples_per_second": 6519.677,
+    "train_steps_per_second": 25.481
+}

trainer_state.json

@@ -0,0 +1,196 @@
+{
+  "best_global_step": 4266,
+  "best_metric": 0.40021559596061707,
+  "best_model_checkpoint": "distilbert_km_50_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": 1.7321940660476685,
+      "learning_rate": 4.9e-05,
+      "loss": 0.4934,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7875340093989611,
+      "eval_combined_score": 0.7305459056647462,
+      "eval_f1": 0.6735578019305313,
+      "eval_loss": 0.4472145140171051,
+      "eval_runtime": 17.1686,
+      "eval_samples_per_second": 2354.88,
+      "eval_steps_per_second": 9.203,
+      "step": 1422
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.5352046489715576,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3917,
+      "step": 2844
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.807395498392283,
+      "eval_combined_score": 0.7771251410234553,
+      "eval_f1": 0.7468547836546277,
+      "eval_loss": 0.4065588414669037,
+      "eval_runtime": 16.7633,
+      "eval_samples_per_second": 2411.812,
+      "eval_steps_per_second": 9.425,
+      "step": 2844
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.6184964179992676,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3258,
+      "step": 4266
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8210734603017561,
+      "eval_combined_score": 0.7933755484976,
+      "eval_f1": 0.7656776366934439,
+      "eval_loss": 0.40021559596061707,
+      "eval_runtime": 16.7318,
+      "eval_samples_per_second": 2416.357,
+      "eval_steps_per_second": 9.443,
+      "step": 4266
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 4.1061296463012695,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.2714,
+      "step": 5688
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8259460796438288,
+      "eval_combined_score": 0.7863946320800718,
+      "eval_f1": 0.7468431845163147,
+      "eval_loss": 0.44828683137893677,
+      "eval_runtime": 16.7387,
+      "eval_samples_per_second": 2415.364,
+      "eval_steps_per_second": 9.439,
+      "step": 5688
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 3.0424695014953613,
+      "learning_rate": 4.5e-05,
+      "loss": 0.2229,
+      "step": 7110
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8319069997526589,
+      "eval_combined_score": 0.7971838389186223,
+      "eval_f1": 0.7624606780845858,
+      "eval_loss": 0.427327424287796,
+      "eval_runtime": 16.3625,
+      "eval_samples_per_second": 2470.89,
+      "eval_steps_per_second": 9.656,
+      "step": 7110
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 3.228925943374634,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1825,
+      "step": 8532
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8314865199109572,
+      "eval_combined_score": 0.8023708766892623,
+      "eval_f1": 0.7732552334675675,
+      "eval_loss": 0.4705049693584442,
+      "eval_runtime": 16.3004,
+      "eval_samples_per_second": 2480.311,
+      "eval_steps_per_second": 9.693,
+      "step": 8532
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 6.286579132080078,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1499,
+      "step": 9954
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8341578036111799,
+      "eval_combined_score": 0.8058221401157482,
+      "eval_f1": 0.7774864766203166,
+      "eval_loss": 0.538841724395752,
+      "eval_runtime": 16.3083,
+      "eval_samples_per_second": 2479.109,
+      "eval_steps_per_second": 9.688,
+      "step": 9954
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.780695915222168,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1219,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8175364828097947,
+      "eval_combined_score": 0.7959942174956425,
+      "eval_f1": 0.7744519521814902,
+      "eval_loss": 0.5893063545227051,
+      "eval_runtime": 16.3197,
+      "eval_samples_per_second": 2477.38,
+      "eval_steps_per_second": 9.682,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "step": 11376,
+      "total_flos": 1.927909325255639e+17,
+      "train_loss": 0.269940188497766,
+      "train_runtime": 2790.3681,
+      "train_samples_per_second": 6519.677,
+      "train_steps_per_second": 25.481
+    }
+  ],
+  "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": 1.927909325255639e+17,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}