End of training

README.md

@@ -1,14 +1,32 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/tiny_bert_km_50_v2
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 - f1
 model-index:
 - name: tiny_bert_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: 0.8069997526589167
+    - name: F1
+      type: f1
+      value: 0.7304197616168595
 ---
 
 <!-- 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_50_v2_qqp
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_km_50_v2](https://huggingface.co/Hartunka/tiny_bert_km_50_v2) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_km_50_v2](https://huggingface.co/Hartunka/tiny_bert_km_50_v2) on the GLUE QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5383
-- Accuracy: 0.8219
-- F1: 0.7543
-- Combined Score: 0.7881
+- Loss: 0.4256
+- Accuracy: 0.8070
+- F1: 0.7304
+- Combined Score: 0.7687
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8069997526589167,
+    "eval_combined_score": 0.768709757137888,
+    "eval_f1": 0.7304197616168595,
+    "eval_loss": 0.4255537688732147,
+    "eval_runtime": 11.4999,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 3515.697,
+    "eval_steps_per_second": 13.739,
+    "total_flos": 7.633075201391002e+16,
+    "train_loss": 0.3116523619274885,
+    "train_runtime": 1602.393,
+    "train_samples": 363846,
+    "train_samples_per_second": 11353.208,
+    "train_steps_per_second": 44.371
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8069997526589167,
+    "eval_combined_score": 0.768709757137888,
+    "eval_f1": 0.7304197616168595,
+    "eval_loss": 0.4255537688732147,
+    "eval_runtime": 11.4999,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 3515.697,
+    "eval_steps_per_second": 13.739
+}

logs/events.out.tfevents.1745285547.s_005_m.2836625.61

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5bd58a6be78b73e46be102d12f5108e33d0200b322392e69154a7840f805fd1
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "total_flos": 7.633075201391002e+16,
+    "train_loss": 0.3116523619274885,
+    "train_runtime": 1602.393,
+    "train_samples": 363846,
+    "train_samples_per_second": 11353.208,
+    "train_steps_per_second": 44.371
+}

trainer_state.json

@@ -0,0 +1,196 @@
+{
+  "best_global_step": 4266,
+  "best_metric": 0.4255537688732147,
+  "best_model_checkpoint": "tiny_bert_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": 1.6206313371658325,
+      "learning_rate": 4.9e-05,
+      "loss": 0.4993,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7789760079149146,
+      "eval_combined_score": 0.7202840351624864,
+      "eval_f1": 0.6615920624100583,
+      "eval_loss": 0.46119508147239685,
+      "eval_runtime": 11.7696,
+      "eval_samples_per_second": 3435.112,
+      "eval_steps_per_second": 13.424,
+      "step": 1422
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.0556938648223877,
+      "learning_rate": 4.8e-05,
+      "loss": 0.4141,
+      "step": 2844
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7986396240415533,
+      "eval_combined_score": 0.7553790786143619,
+      "eval_f1": 0.7121185331871707,
+      "eval_loss": 0.427831768989563,
+      "eval_runtime": 11.4532,
+      "eval_samples_per_second": 3530.019,
+      "eval_steps_per_second": 13.795,
+      "step": 2844
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.7704219818115234,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3569,
+      "step": 4266
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8069997526589167,
+      "eval_combined_score": 0.768709757137888,
+      "eval_f1": 0.7304197616168595,
+      "eval_loss": 0.4255537688732147,
+      "eval_runtime": 11.4748,
+      "eval_samples_per_second": 3523.368,
+      "eval_steps_per_second": 13.769,
+      "step": 4266
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.501729726791382,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.3108,
+      "step": 5688
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8137274301261439,
+      "eval_combined_score": 0.7731789065992668,
+      "eval_f1": 0.7326303830723897,
+      "eval_loss": 0.4395003616809845,
+      "eval_runtime": 11.476,
+      "eval_samples_per_second": 3522.99,
+      "eval_steps_per_second": 13.768,
+      "step": 5688
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 3.398540735244751,
+      "learning_rate": 4.5e-05,
+      "loss": 0.273,
+      "step": 7110
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.817116002968093,
+      "eval_combined_score": 0.7853246681507131,
+      "eval_f1": 0.7535333333333334,
+      "eval_loss": 0.4446704685688019,
+      "eval_runtime": 11.4901,
+      "eval_samples_per_second": 3518.693,
+      "eval_steps_per_second": 13.751,
+      "step": 7110
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.8529112339019775,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.2403,
+      "step": 8532
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8186742517932228,
+      "eval_combined_score": 0.7888684631957208,
+      "eval_f1": 0.7590626745982186,
+      "eval_loss": 0.4628998935222626,
+      "eval_runtime": 11.4817,
+      "eval_samples_per_second": 3521.268,
+      "eval_steps_per_second": 13.761,
+      "step": 8532
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 4.417098522186279,
+      "learning_rate": 4.3e-05,
+      "loss": 0.2117,
+      "step": 9954
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8171902052930992,
+      "eval_combined_score": 0.7867881874548821,
+      "eval_f1": 0.756386169616665,
+      "eval_loss": 0.4907230734825134,
+      "eval_runtime": 11.4595,
+      "eval_samples_per_second": 3528.079,
+      "eval_steps_per_second": 13.788,
+      "step": 9954
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 4.252416610717773,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1871,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8219144199851596,
+      "eval_combined_score": 0.7880987020409042,
+      "eval_f1": 0.7542829840966487,
+      "eval_loss": 0.5383203029632568,
+      "eval_runtime": 11.442,
+      "eval_samples_per_second": 3533.487,
+      "eval_steps_per_second": 13.809,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "step": 11376,
+      "total_flos": 7.633075201391002e+16,
+      "train_loss": 0.3116523619274885,
+      "train_runtime": 1602.393,
+      "train_samples_per_second": 11353.208,
+      "train_steps_per_second": 44.371
+    }
+  ],
+  "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
+}