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_mrpc
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE MRPC
+      type: glue
+      args: mrpc
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.7181372549019608
+    - name: F1
+      type: f1
+      value: 0.8211508553654744
 ---
 
 <!-- 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_mrpc
 
-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 MRPC dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8555
-- Accuracy: 0.7034
-- F1: 0.8
-- Combined Score: 0.7517
+- Loss: 0.5812
+- Accuracy: 0.7181
+- F1: 0.8212
+- Combined Score: 0.7696
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.7181372549019608,
+    "eval_combined_score": 0.7696440551337176,
+    "eval_f1": 0.8211508553654744,
+    "eval_loss": 0.5812366008758545,
+    "eval_runtime": 0.1548,
+    "eval_samples": 408,
+    "eval_samples_per_second": 2636.086,
+    "eval_steps_per_second": 12.922,
+    "total_flos": 865692760633344.0,
+    "train_loss": 0.45122435004622846,
+    "train_runtime": 25.3703,
+    "train_samples": 3668,
+    "train_samples_per_second": 7228.916,
+    "train_steps_per_second": 29.562
+}

eval_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 9.0,
+    "eval_accuracy": 0.7181372549019608,
+    "eval_combined_score": 0.7696440551337176,
+    "eval_f1": 0.8211508553654744,
+    "eval_loss": 0.5812366008758545,
+    "eval_runtime": 0.1548,
+    "eval_samples": 408,
+    "eval_samples_per_second": 2636.086,
+    "eval_steps_per_second": 12.922
+}

logs/events.out.tfevents.1745283357.s_005_m.2836625.57

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ee4455c1c7ad4d60b2f42ff7fdc8959fd5769e6317566f76233cc55c0de32a6
+size 467

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 9.0,
+    "total_flos": 865692760633344.0,
+    "train_loss": 0.45122435004622846,
+    "train_runtime": 25.3703,
+    "train_samples": 3668,
+    "train_samples_per_second": 7228.916,
+    "train_steps_per_second": 29.562
+}

trainer_state.json

@@ -0,0 +1,214 @@
+{
+  "best_global_step": 60,
+  "best_metric": 0.5812366008758545,
+  "best_model_checkpoint": "tiny_bert_km_50_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": 0.7363313436508179,
+      "learning_rate": 4.9e-05,
+      "loss": 0.632,
+      "step": 15
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7034313725490197,
+      "eval_combined_score": 0.7601876983773843,
+      "eval_f1": 0.8169440242057489,
+      "eval_loss": 0.5975198149681091,
+      "eval_runtime": 0.1327,
+      "eval_samples_per_second": 3075.467,
+      "eval_steps_per_second": 15.076,
+      "step": 15
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.268504023551941,
+      "learning_rate": 4.8e-05,
+      "loss": 0.5917,
+      "step": 30
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7132352941176471,
+      "eval_combined_score": 0.7691737008704828,
+      "eval_f1": 0.8251121076233184,
+      "eval_loss": 0.5832136273384094,
+      "eval_runtime": 0.1486,
+      "eval_samples_per_second": 2746.391,
+      "eval_steps_per_second": 13.463,
+      "step": 30
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.0295687913894653,
+      "learning_rate": 4.7e-05,
+      "loss": 0.5638,
+      "step": 45
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7034313725490197,
+      "eval_combined_score": 0.7607382426654873,
+      "eval_f1": 0.8180451127819549,
+      "eval_loss": 0.5895337462425232,
+      "eval_runtime": 0.1302,
+      "eval_samples_per_second": 3134.825,
+      "eval_steps_per_second": 15.367,
+      "step": 45
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.3483902215957642,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.5417,
+      "step": 60
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7181372549019608,
+      "eval_combined_score": 0.7696440551337176,
+      "eval_f1": 0.8211508553654744,
+      "eval_loss": 0.5812366008758545,
+      "eval_runtime": 0.1387,
+      "eval_samples_per_second": 2941.894,
+      "eval_steps_per_second": 14.421,
+      "step": 60
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.3638129234313965,
+      "learning_rate": 4.5e-05,
+      "loss": 0.5019,
+      "step": 75
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.7009803921568627,
+      "eval_combined_score": 0.7524194565285921,
+      "eval_f1": 0.8038585209003215,
+      "eval_loss": 0.5993048548698425,
+      "eval_runtime": 0.1529,
+      "eval_samples_per_second": 2667.691,
+      "eval_steps_per_second": 13.077,
+      "step": 75
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.534933090209961,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.4343,
+      "step": 90
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.678921568627451,
+      "eval_combined_score": 0.7301118694556287,
+      "eval_f1": 0.7813021702838063,
+      "eval_loss": 0.620622456073761,
+      "eval_runtime": 0.1423,
+      "eval_samples_per_second": 2867.218,
+      "eval_steps_per_second": 14.055,
+      "step": 90
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.426776885986328,
+      "learning_rate": 4.3e-05,
+      "loss": 0.3592,
+      "step": 105
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7083333333333334,
+      "eval_combined_score": 0.754501675041876,
+      "eval_f1": 0.8006700167504187,
+      "eval_loss": 0.6731041669845581,
+      "eval_runtime": 0.1344,
+      "eval_samples_per_second": 3035.9,
+      "eval_steps_per_second": 14.882,
+      "step": 105
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.3238067626953125,
+      "learning_rate": 4.2e-05,
+      "loss": 0.2533,
+      "step": 120
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7009803921568627,
+      "eval_combined_score": 0.7508169934640523,
+      "eval_f1": 0.8006535947712419,
+      "eval_loss": 0.7867075800895691,
+      "eval_runtime": 0.1325,
+      "eval_samples_per_second": 3079.546,
+      "eval_steps_per_second": 15.096,
+      "step": 120
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.622627019882202,
+      "learning_rate": 4.1e-05,
+      "loss": 0.1831,
+      "step": 135
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.7034313725490197,
+      "eval_combined_score": 0.7517156862745098,
+      "eval_f1": 0.8,
+      "eval_loss": 0.8554521203041077,
+      "eval_runtime": 0.1556,
+      "eval_samples_per_second": 2622.352,
+      "eval_steps_per_second": 12.855,
+      "step": 135
+    },
+    {
+      "epoch": 9.0,
+      "step": 135,
+      "total_flos": 865692760633344.0,
+      "train_loss": 0.45122435004622846,
+      "train_runtime": 25.3703,
+      "train_samples_per_second": 7228.916,
+      "train_steps_per_second": 29.562
+    }
+  ],
+  "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
+}