End of training

README.md

@@ -1,14 +1,32 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/distilbert_km_20_v1
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 - f1
 model-index:
 - name: distilbert_km_20_v1_qqp
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE QQP
+      type: glue
+      args: qqp
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.8151867425179322
+    - name: F1
+      type: f1
+      value: 0.7475846226606311
 ---
 
 <!-- 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_20_v1_qqp
 
-This model is a fine-tuned version of [Hartunka/distilbert_km_20_v1](https://huggingface.co/Hartunka/distilbert_km_20_v1) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/distilbert_km_20_v1](https://huggingface.co/Hartunka/distilbert_km_20_v1) on the GLUE QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5558
-- Accuracy: 0.8324
-- F1: 0.7772
-- Combined Score: 0.8048
+- Loss: 0.3983
+- Accuracy: 0.8152
+- F1: 0.7476
+- Combined Score: 0.7814
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.8151867425179322,
+    "eval_combined_score": 0.7813856825892816,
+    "eval_f1": 0.7475846226606311,
+    "eval_loss": 0.39828023314476013,
+    "eval_runtime": 16.2085,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 2494.369,
+    "eval_steps_per_second": 9.748,
+    "total_flos": 1.686920659598684e+17,
+    "train_loss": 0.27118523896447094,
+    "train_runtime": 2421.472,
+    "train_samples": 363846,
+    "train_samples_per_second": 7512.909,
+    "train_steps_per_second": 29.362
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 7.0,
+    "eval_accuracy": 0.8151867425179322,
+    "eval_combined_score": 0.7813856825892816,
+    "eval_f1": 0.7475846226606311,
+    "eval_loss": 0.39828023314476013,
+    "eval_runtime": 16.2085,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 2494.369,
+    "eval_steps_per_second": 9.748
+}

logs/events.out.tfevents.1744681642.s_005_m.602156.43

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae919fe1d62a4385fcd8613256148c38d555a84ac41c0c6d2a55effe6658bf71
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 7.0,
+    "total_flos": 1.686920659598684e+17,
+    "train_loss": 0.27118523896447094,
+    "train_runtime": 2421.472,
+    "train_samples": 363846,
+    "train_samples_per_second": 7512.909,
+    "train_steps_per_second": 29.362
+}

trainer_state.json

@@ -0,0 +1,178 @@
+{
+  "best_global_step": 2844,
+  "best_metric": 0.39828023314476013,
+  "best_model_checkpoint": "distilbert_km_20_v1_qqp/checkpoint-2844",
+  "epoch": 7.0,
+  "eval_steps": 500,
+  "global_step": 9954,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7204290628433228,
+      "learning_rate": 4.9e-05,
+      "loss": 0.485,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7906010388325501,
+      "eval_combined_score": 0.7413396552277248,
+      "eval_f1": 0.6920782716228996,
+      "eval_loss": 0.43430814146995544,
+      "eval_runtime": 16.503,
+      "eval_samples_per_second": 2449.86,
+      "eval_steps_per_second": 9.574,
+      "step": 1422
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.371051549911499,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3809,
+      "step": 2844
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8151867425179322,
+      "eval_combined_score": 0.7813856825892816,
+      "eval_f1": 0.7475846226606311,
+      "eval_loss": 0.39828023314476013,
+      "eval_runtime": 16.3048,
+      "eval_samples_per_second": 2479.641,
+      "eval_steps_per_second": 9.69,
+      "step": 2844
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.6035635471343994,
+      "learning_rate": 4.7e-05,
+      "loss": 0.307,
+      "step": 4266
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8230274548602523,
+      "eval_combined_score": 0.7956582916971038,
+      "eval_f1": 0.7682891285339551,
+      "eval_loss": 0.3996948301792145,
+      "eval_runtime": 16.2668,
+      "eval_samples_per_second": 2485.435,
+      "eval_steps_per_second": 9.713,
+      "step": 4266
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.597846508026123,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.2452,
+      "step": 5688
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8293099183774425,
+      "eval_combined_score": 0.7927162377694895,
+      "eval_f1": 0.7561225571615365,
+      "eval_loss": 0.43759506940841675,
+      "eval_runtime": 16.3498,
+      "eval_samples_per_second": 2472.818,
+      "eval_steps_per_second": 9.664,
+      "step": 5688
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 3.629729986190796,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1961,
+      "step": 7110
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8329458323027454,
+      "eval_combined_score": 0.8000647810255541,
+      "eval_f1": 0.7671837297483627,
+      "eval_loss": 0.4591920077800751,
+      "eval_runtime": 16.3052,
+      "eval_samples_per_second": 2479.577,
+      "eval_steps_per_second": 9.69,
+      "step": 7110
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 4.918776512145996,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1568,
+      "step": 8532
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8239673509769972,
+      "eval_combined_score": 0.7989472408136153,
+      "eval_f1": 0.7739271306502334,
+      "eval_loss": 0.5181909799575806,
+      "eval_runtime": 16.2847,
+      "eval_samples_per_second": 2482.694,
+      "eval_steps_per_second": 9.702,
+      "step": 8532
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 5.020405292510986,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1273,
+      "step": 9954
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8324016819193668,
+      "eval_combined_score": 0.8047827930775471,
+      "eval_f1": 0.7771639042357275,
+      "eval_loss": 0.5558242797851562,
+      "eval_runtime": 16.3405,
+      "eval_samples_per_second": 2474.224,
+      "eval_steps_per_second": 9.669,
+      "step": 9954
+    },
+    {
+      "epoch": 7.0,
+      "step": 9954,
+      "total_flos": 1.686920659598684e+17,
+      "train_loss": 0.27118523896447094,
+      "train_runtime": 2421.472,
+      "train_samples_per_second": 7512.909,
+      "train_steps_per_second": 29.362
+    }
+  ],
+  "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": 1.686920659598684e+17,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}