End of training

README.md

@@ -1,14 +1,32 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/tiny_bert_rand_20_v1
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 - f1
 model-index:
 - name: tiny_bert_rand_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: 0.8180311649765026
+    - name: F1
+      type: f1
+      value: 0.7436138700121973
 ---
 
 <!-- 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_rand_20_v1_qqp
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_rand_20_v1](https://huggingface.co/Hartunka/tiny_bert_rand_20_v1) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_rand_20_v1](https://huggingface.co/Hartunka/tiny_bert_rand_20_v1) on the GLUE QQP dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5970
-- Accuracy: 0.8317
-- F1: 0.7599
-- Combined Score: 0.7958
+- Loss: 0.4170
+- Accuracy: 0.8180
+- F1: 0.7436
+- Combined Score: 0.7808
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8180311649765026,
+    "eval_combined_score": 0.7808225174943499,
+    "eval_f1": 0.7436138700121973,
+    "eval_loss": 0.41695910692214966,
+    "eval_runtime": 11.8805,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 3403.06,
+    "eval_steps_per_second": 13.299,
+    "total_flos": 7.633075201391002e+16,
+    "train_loss": 0.2682548877056138,
+    "train_runtime": 1625.0507,
+    "train_samples": 363846,
+    "train_samples_per_second": 11194.912,
+    "train_steps_per_second": 43.752
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.8180311649765026,
+    "eval_combined_score": 0.7808225174943499,
+    "eval_f1": 0.7436138700121973,
+    "eval_loss": 0.41695910692214966,
+    "eval_runtime": 11.8805,
+    "eval_samples": 40430,
+    "eval_samples_per_second": 3403.06,
+    "eval_steps_per_second": 13.299
+}

logs/events.out.tfevents.1744654423.s_005_m.2701049.97

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07d9033c2880b0185d403a58cca8497f40a97ede75ad65fda5a6015cee157ecc
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "total_flos": 7.633075201391002e+16,
+    "train_loss": 0.2682548877056138,
+    "train_runtime": 1625.0507,
+    "train_samples": 363846,
+    "train_samples_per_second": 11194.912,
+    "train_steps_per_second": 43.752
+}

trainer_state.json

@@ -0,0 +1,196 @@
+{
+  "best_global_step": 4266,
+  "best_metric": 0.41695910692214966,
+  "best_model_checkpoint": "tiny_bert_rand_20_v1_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": 2.032697916030884,
+      "learning_rate": 4.9e-05,
+      "loss": 0.4821,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7931981202077665,
+      "eval_combined_score": 0.7460317991638508,
+      "eval_f1": 0.6988654781199352,
+      "eval_loss": 0.43809574842453003,
+      "eval_runtime": 12.0601,
+      "eval_samples_per_second": 3352.379,
+      "eval_steps_per_second": 13.101,
+      "step": 1422
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.111968517303467,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3837,
+      "step": 2844
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8095473658174622,
+      "eval_combined_score": 0.7618316765227506,
+      "eval_f1": 0.7141159872280389,
+      "eval_loss": 0.4237190783023834,
+      "eval_runtime": 11.8112,
+      "eval_samples_per_second": 3423.025,
+      "eval_steps_per_second": 13.377,
+      "step": 2844
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.4837989807128906,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3187,
+      "step": 4266
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8180311649765026,
+      "eval_combined_score": 0.7808225174943499,
+      "eval_f1": 0.7436138700121973,
+      "eval_loss": 0.41695910692214966,
+      "eval_runtime": 11.7905,
+      "eval_samples_per_second": 3429.039,
+      "eval_steps_per_second": 13.401,
+      "step": 4266
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.610124349594116,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.2638,
+      "step": 5688
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8243630967103636,
+      "eval_combined_score": 0.7880077995355406,
+      "eval_f1": 0.7516525023607177,
+      "eval_loss": 0.43834444880485535,
+      "eval_runtime": 11.7911,
+      "eval_samples_per_second": 3428.855,
+      "eval_steps_per_second": 13.4,
+      "step": 5688
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 3.3936569690704346,
+      "learning_rate": 4.5e-05,
+      "loss": 0.2202,
+      "step": 7110
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8277764036606481,
+      "eval_combined_score": 0.7926280576347196,
+      "eval_f1": 0.757479711608791,
+      "eval_loss": 0.46568045020103455,
+      "eval_runtime": 11.5465,
+      "eval_samples_per_second": 3501.494,
+      "eval_steps_per_second": 13.684,
+      "step": 7110
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 5.742647647857666,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1852,
+      "step": 8532
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8259213455354935,
+      "eval_combined_score": 0.7949921846685177,
+      "eval_f1": 0.7640630238015421,
+      "eval_loss": 0.5005447864532471,
+      "eval_runtime": 11.3692,
+      "eval_samples_per_second": 3556.099,
+      "eval_steps_per_second": 13.897,
+      "step": 8532
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 4.802578449249268,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1567,
+      "step": 9954
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8237447440019787,
+      "eval_combined_score": 0.7951369025035777,
+      "eval_f1": 0.7665290610051766,
+      "eval_loss": 0.5466973185539246,
+      "eval_runtime": 11.7132,
+      "eval_samples_per_second": 3451.654,
+      "eval_steps_per_second": 13.489,
+      "step": 9954
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.240032434463501,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1357,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8316843927776404,
+      "eval_combined_score": 0.7958118580814155,
+      "eval_f1": 0.7599393233851907,
+      "eval_loss": 0.5970303416252136,
+      "eval_runtime": 11.8276,
+      "eval_samples_per_second": 3418.281,
+      "eval_steps_per_second": 13.359,
+      "step": 11376
+    },
+    {
+      "epoch": 8.0,
+      "step": 11376,
+      "total_flos": 7.633075201391002e+16,
+      "train_loss": 0.2682548877056138,
+      "train_runtime": 1625.0507,
+      "train_samples_per_second": 11194.912,
+      "train_steps_per_second": 43.752
+    }
+  ],
+  "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
+}