End of training

README.md

@@ -1,14 +1,32 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/tiny_bert_km_5_v2
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 - f1
 model-index:
 - name: tiny_bert_km_5_v2_mrpc
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE MRPC
+      type: glue
+      args: mrpc
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.6838235294117647
+    - name: F1
+      type: f1
+      value: 0.7817258883248731
 ---
 
 <!-- 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_5_v2_mrpc
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_km_5_v2](https://huggingface.co/Hartunka/tiny_bert_km_5_v2) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_km_5_v2](https://huggingface.co/Hartunka/tiny_bert_km_5_v2) on the GLUE MRPC dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8383
-- Accuracy: 0.6299
-- F1: 0.7299
-- Combined Score: 0.6799
+- Loss: 0.5942
+- Accuracy: 0.6838
+- F1: 0.7817
+- Combined Score: 0.7328
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.6838235294117647,
+    "eval_combined_score": 0.732774708868319,
+    "eval_f1": 0.7817258883248731,
+    "eval_loss": 0.5941939949989319,
+    "eval_runtime": 0.1642,
+    "eval_samples": 408,
+    "eval_samples_per_second": 2484.741,
+    "eval_steps_per_second": 12.18,
+    "total_flos": 865692760633344.0,
+    "train_loss": 0.4928195600156431,
+    "train_runtime": 26.9972,
+    "train_samples": 3668,
+    "train_samples_per_second": 6793.288,
+    "train_steps_per_second": 27.781
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.6838235294117647,
+    "eval_combined_score": 0.732774708868319,
+    "eval_f1": 0.7817258883248731,
+    "eval_loss": 0.5941939949989319,
+    "eval_runtime": 0.1642,
+    "eval_samples": 408,
+    "eval_samples_per_second": 2484.741,
+    "eval_steps_per_second": 12.18
+}

logs/events.out.tfevents.1745268076.s_005_m.2836625.3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c8301e5c18dd55b731a4874b1777fee03a22a2a3eb804d459e2533e5e29032c
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 9.0,
+    "total_flos": 865692760633344.0,
+    "train_loss": 0.4928195600156431,
+    "train_runtime": 26.9972,
+    "train_samples": 3668,
+    "train_samples_per_second": 6793.288,
+    "train_steps_per_second": 27.781
+}

trainer_state.json

@@ -0,0 +1,214 @@
+{
+  "best_global_step": 60,
+  "best_metric": 0.5941939949989319,
+  "best_model_checkpoint": "tiny_bert_km_5_v2_mrpc/checkpoint-60",
+  "epoch": 9.0,
+  "eval_steps": 500,
+  "global_step": 135,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.790999710559845,
+      "learning_rate": 4.9e-05,
+      "loss": 0.6277,
+      "step": 15
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.6887254901960784,
+      "eval_combined_score": 0.7491603462974394,
+      "eval_f1": 0.8095952023988006,
+      "eval_loss": 0.6063878536224365,
+      "eval_runtime": 0.1348,
+      "eval_samples_per_second": 3027.676,
+      "eval_steps_per_second": 14.842,
+      "step": 15
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.333164095878601,
+      "learning_rate": 4.8e-05,
+      "loss": 0.5964,
+      "step": 30
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6764705882352942,
+      "eval_combined_score": 0.73109243697479,
+      "eval_f1": 0.7857142857142857,
+      "eval_loss": 0.600712776184082,
+      "eval_runtime": 0.1473,
+      "eval_samples_per_second": 2770.333,
+      "eval_steps_per_second": 13.58,
+      "step": 30
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.0239461660385132,
+      "learning_rate": 4.7e-05,
+      "loss": 0.5743,
+      "step": 45
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6985294117647058,
+      "eval_combined_score": 0.7570608078313784,
+      "eval_f1": 0.815592203898051,
+      "eval_loss": 0.6121370792388916,
+      "eval_runtime": 0.1391,
+      "eval_samples_per_second": 2933.075,
+      "eval_steps_per_second": 14.378,
+      "step": 45
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.6916087865829468,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.5591,
+      "step": 60
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6838235294117647,
+      "eval_combined_score": 0.732774708868319,
+      "eval_f1": 0.7817258883248731,
+      "eval_loss": 0.5941939949989319,
+      "eval_runtime": 0.1743,
+      "eval_samples_per_second": 2341.276,
+      "eval_steps_per_second": 11.477,
+      "step": 60
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 2.1864917278289795,
+      "learning_rate": 4.5e-05,
+      "loss": 0.5162,
+      "step": 75
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6813725490196079,
+      "eval_combined_score": 0.7358475648323846,
+      "eval_f1": 0.7903225806451613,
+      "eval_loss": 0.6068432331085205,
+      "eval_runtime": 0.1684,
+      "eval_samples_per_second": 2422.792,
+      "eval_steps_per_second": 11.876,
+      "step": 75
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.5886662006378174,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.4733,
+      "step": 90
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6544117647058824,
+      "eval_combined_score": 0.6964081086980247,
+      "eval_f1": 0.738404452690167,
+      "eval_loss": 0.6536080837249756,
+      "eval_runtime": 0.1316,
+      "eval_samples_per_second": 3099.593,
+      "eval_steps_per_second": 15.194,
+      "step": 90
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 3.5240910053253174,
+      "learning_rate": 4.3e-05,
+      "loss": 0.4261,
+      "step": 105
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7058823529411765,
+      "eval_combined_score": 0.7539312754804892,
+      "eval_f1": 0.801980198019802,
+      "eval_loss": 0.6742849946022034,
+      "eval_runtime": 0.1506,
+      "eval_samples_per_second": 2708.667,
+      "eval_steps_per_second": 13.278,
+      "step": 105
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 4.345057010650635,
+      "learning_rate": 4.2e-05,
+      "loss": 0.3647,
+      "step": 120
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6813725490196079,
+      "eval_combined_score": 0.7293849046467902,
+      "eval_f1": 0.7773972602739726,
+      "eval_loss": 0.7378984093666077,
+      "eval_runtime": 0.1633,
+      "eval_samples_per_second": 2498.228,
+      "eval_steps_per_second": 12.246,
+      "step": 120
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 6.365061283111572,
+      "learning_rate": 4.1e-05,
+      "loss": 0.2976,
+      "step": 135
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6299019607843137,
+      "eval_combined_score": 0.679888368585359,
+      "eval_f1": 0.7298747763864043,
+      "eval_loss": 0.8383299112319946,
+      "eval_runtime": 0.1418,
+      "eval_samples_per_second": 2876.476,
+      "eval_steps_per_second": 14.1,
+      "step": 135
+    },
+    {
+      "epoch": 9.0,
+      "step": 135,
+      "total_flos": 865692760633344.0,
+      "train_loss": 0.4928195600156431,
+      "train_runtime": 26.9972,
+      "train_samples_per_second": 6793.288,
+      "train_steps_per_second": 27.781
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 750,
+  "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": 865692760633344.0,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}