End of training

README.md

@@ -1,14 +1,32 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/distilbert_km_100_v2
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 - f1
 model-index:
 - name: distilbert_km_100_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: 0.8217907494434825
+    - name: F1
+      type: f1
+      value: 0.758505111446288
 ---
 
 <!-- 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_100_v2_qqp
 
-This model is a fine-tuned version of [Hartunka/distilbert_km_100_v2](https://huggingface.co/Hartunka/distilbert_km_100_v2) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/distilbert_km_100_v2](https://huggingface.co/Hartunka/distilbert_km_100_v2) on the GLUE QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6220
-- Accuracy: 0.8306
-- F1: 0.7703
-- Combined Score: 0.8005
+- Loss: 0.4029
+- Accuracy: 0.8218
+- F1: 0.7585
+- Combined Score: 0.7901
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8217907494434825,
+    "eval_combined_score": 0.7901479304448853,
+    "eval_f1": 0.758505111446288,
+    "eval_loss": 0.4029361605644226,
+    "eval_runtime": 16.2788,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 2483.599,
+    "eval_steps_per_second": 9.706,
+    "total_flos": 1.927909325255639e+17,
+    "train_loss": 0.25120435653021234,
+    "train_runtime": 2768.7641,
+    "train_samples": 363846,
+    "train_samples_per_second": 6570.549,
+    "train_steps_per_second": 25.679
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8217907494434825,
+    "eval_combined_score": 0.7901479304448853,
+    "eval_f1": 0.758505111446288,
+    "eval_loss": 0.4029361605644226,
+    "eval_runtime": 16.2788,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 2483.599,
+    "eval_steps_per_second": 9.706
+}

logs/events.out.tfevents.1745342007.s_005_m.2850345.79

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3ab230f612921ca5178035bed4cf3b39ff7b0cb58d28479bdf1908f81cdb9ee
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "total_flos": 1.927909325255639e+17,
+    "train_loss": 0.25120435653021234,
+    "train_runtime": 2768.7641,
+    "train_samples": 363846,
+    "train_samples_per_second": 6570.549,
+    "train_steps_per_second": 25.679
+}

trainer_state.json

@@ -0,0 +1,196 @@
+{
+  "best_global_step": 4266,
+  "best_metric": 0.4029361605644226,
+  "best_model_checkpoint": "distilbert_km_100_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": 1.6681804656982422,
+      "learning_rate": 4.9e-05,
+      "loss": 0.4895,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7879050210239921,
+      "eval_combined_score": 0.7424132605208321,
+      "eval_f1": 0.6969215000176722,
+      "eval_loss": 0.4377460479736328,
+      "eval_runtime": 16.5991,
+      "eval_samples_per_second": 2435.673,
+      "eval_steps_per_second": 9.519,
+      "step": 1422
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.6809704303741455,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3864,
+      "step": 2844
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8124659906010389,
+      "eval_combined_score": 0.7768208289865985,
+      "eval_f1": 0.7411756673721581,
+      "eval_loss": 0.40619000792503357,
+      "eval_runtime": 16.2634,
+      "eval_samples_per_second": 2485.957,
+      "eval_steps_per_second": 9.715,
+      "step": 2844
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.2454617023468018,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3094,
+      "step": 4266
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8217907494434825,
+      "eval_combined_score": 0.7901479304448853,
+      "eval_f1": 0.758505111446288,
+      "eval_loss": 0.4029361605644226,
+      "eval_runtime": 16.3662,
+      "eval_samples_per_second": 2470.335,
+      "eval_steps_per_second": 9.654,
+      "step": 4266
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 4.337860107421875,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.2466,
+      "step": 5688
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8274053920356171,
+      "eval_combined_score": 0.7908765325312213,
+      "eval_f1": 0.7543476730268254,
+      "eval_loss": 0.4315114915370941,
+      "eval_runtime": 16.3558,
+      "eval_samples_per_second": 2471.899,
+      "eval_steps_per_second": 9.66,
+      "step": 5688
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 3.1645774841308594,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1952,
+      "step": 7110
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8261934207271828,
+      "eval_combined_score": 0.7969056987560648,
+      "eval_f1": 0.7676179767849466,
+      "eval_loss": 0.4505920112133026,
+      "eval_runtime": 16.3863,
+      "eval_samples_per_second": 2467.303,
+      "eval_steps_per_second": 9.642,
+      "step": 7110
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 4.501672744750977,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1554,
+      "step": 8532
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8310907741775909,
+      "eval_combined_score": 0.8012830974927041,
+      "eval_f1": 0.7714754208078172,
+      "eval_loss": 0.5191759467124939,
+      "eval_runtime": 16.2852,
+      "eval_samples_per_second": 2482.626,
+      "eval_steps_per_second": 9.702,
+      "step": 8532
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 8.976921081542969,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1248,
+      "step": 9954
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8285926292357161,
+      "eval_combined_score": 0.799057522546553,
+      "eval_f1": 0.7695224158573899,
+      "eval_loss": 0.5748887658119202,
+      "eval_runtime": 16.2532,
+      "eval_samples_per_second": 2487.516,
+      "eval_steps_per_second": 9.721,
+      "step": 9954
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.0511038303375244,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1022,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.830596092010883,
+      "eval_combined_score": 0.8004705638607881,
+      "eval_f1": 0.7703450357106931,
+      "eval_loss": 0.622015655040741,
+      "eval_runtime": 16.275,
+      "eval_samples_per_second": 2484.174,
+      "eval_steps_per_second": 9.708,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "step": 11376,
+      "total_flos": 1.927909325255639e+17,
+      "train_loss": 0.25120435653021234,
+      "train_runtime": 2768.7641,
+      "train_samples_per_second": 6570.549,
+      "train_steps_per_second": 25.679
+    }
+  ],
+  "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
+}