End of training

README.md

@@ -1,13 +1,28 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/distilbert_km_20_v2
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: distilbert_km_20_v2_mnli
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE MNLI
+      type: glue
+      args: mnli
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.6598860862489829
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -15,10 +30,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # distilbert_km_20_v2_mnli
 
-This model is a fine-tuned version of [Hartunka/distilbert_km_20_v2](https://huggingface.co/Hartunka/distilbert_km_20_v2) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/distilbert_km_20_v2](https://huggingface.co/Hartunka/distilbert_km_20_v2) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0935
-- Accuracy: 0.6529
+- Loss: 0.7786
+- Accuracy: 0.6599
 
 ## Model description
 

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 9.0,
+    "epoch_mm": 9.0,
+    "eval_accuracy": 0.656444218033622,
+    "eval_accuracy_mm": 0.6598860862489829,
+    "eval_loss": 0.7839012145996094,
+    "eval_loss_mm": 0.7786339521408081,
+    "eval_runtime": 4.0661,
+    "eval_runtime_mm": 4.082,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2413.85,
+    "eval_samples_per_second_mm": 2408.601,
+    "eval_steps_per_second": 9.591,
+    "eval_steps_per_second_mm": 9.554,
+    "total_flos": 2.340951304214108e+17,
+    "train_loss": 0.6607053365186558,
+    "train_runtime": 3259.3556,
+    "train_samples": 392702,
+    "train_samples_per_second": 6024.228,
+    "train_steps_per_second": 23.532
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 9.0,
+    "epoch_mm": 9.0,
+    "eval_accuracy": 0.656444218033622,
+    "eval_accuracy_mm": 0.6598860862489829,
+    "eval_loss": 0.7839012145996094,
+    "eval_loss_mm": 0.7786339521408081,
+    "eval_runtime": 4.0661,
+    "eval_runtime_mm": 4.082,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2413.85,
+    "eval_samples_per_second_mm": 2408.601,
+    "eval_steps_per_second": 9.591,
+    "eval_steps_per_second_mm": 9.554
+}

logs/events.out.tfevents.1745329991.s_005_m.2850345.53

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c35139910563f7434874ed95884437d8f6b9d60a6fbd1be1042ae49e4da16990
+size 686

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 9.0,
+    "total_flos": 2.340951304214108e+17,
+    "train_loss": 0.6607053365186558,
+    "train_runtime": 3259.3556,
+    "train_samples": 392702,
+    "train_samples_per_second": 6024.228,
+    "train_steps_per_second": 23.532
+}

trainer_state.json

@@ -0,0 +1,196 @@
+{
+  "best_global_step": 6136,
+  "best_metric": 0.7839012145996094,
+  "best_model_checkpoint": "distilbert_km_20_v2_mnli/checkpoint-6136",
+  "epoch": 9.0,
+  "eval_steps": 500,
+  "global_step": 13806,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2019141912460327,
+      "learning_rate": 4.9e-05,
+      "loss": 0.9903,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5645440652063168,
+      "eval_loss": 0.9187681078910828,
+      "eval_runtime": 4.0491,
+      "eval_samples_per_second": 2423.967,
+      "eval_steps_per_second": 9.632,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.564088225364685,
+      "learning_rate": 4.8e-05,
+      "loss": 0.8821,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.6134488028527764,
+      "eval_loss": 0.8513433933258057,
+      "eval_runtime": 3.9821,
+      "eval_samples_per_second": 2464.759,
+      "eval_steps_per_second": 9.794,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.7358378171920776,
+      "learning_rate": 4.7e-05,
+      "loss": 0.791,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6461538461538462,
+      "eval_loss": 0.8013883233070374,
+      "eval_runtime": 3.9811,
+      "eval_samples_per_second": 2465.423,
+      "eval_steps_per_second": 9.796,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.051180601119995,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.7157,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.656444218033622,
+      "eval_loss": 0.7839012145996094,
+      "eval_runtime": 3.9695,
+      "eval_samples_per_second": 2472.591,
+      "eval_steps_per_second": 9.825,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 2.26766300201416,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6472,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.665206316861946,
+      "eval_loss": 0.7938839197158813,
+      "eval_runtime": 4.2948,
+      "eval_samples_per_second": 2285.332,
+      "eval_steps_per_second": 9.081,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.4117183685302734,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.5791,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6662251655629139,
+      "eval_loss": 0.8591964840888977,
+      "eval_runtime": 3.9924,
+      "eval_samples_per_second": 2458.433,
+      "eval_steps_per_second": 9.769,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.8018412590026855,
+      "learning_rate": 4.3e-05,
+      "loss": 0.5111,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6664289353031075,
+      "eval_loss": 0.8661112189292908,
+      "eval_runtime": 3.9981,
+      "eval_samples_per_second": 2454.941,
+      "eval_steps_per_second": 9.755,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 3.3309311866760254,
+      "learning_rate": 4.2e-05,
+      "loss": 0.445,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6538970962812023,
+      "eval_loss": 0.9801768064498901,
+      "eval_runtime": 3.978,
+      "eval_samples_per_second": 2467.348,
+      "eval_steps_per_second": 9.804,
+      "step": 12272
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 4.193544387817383,
+      "learning_rate": 4.1e-05,
+      "loss": 0.3847,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6528782475802344,
+      "eval_loss": 1.0934809446334839,
+      "eval_runtime": 4.0083,
+      "eval_samples_per_second": 2448.671,
+      "eval_steps_per_second": 9.73,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "step": 13806,
+      "total_flos": 2.340951304214108e+17,
+      "train_loss": 0.6607053365186558,
+      "train_runtime": 3259.3556,
+      "train_samples_per_second": 6024.228,
+      "train_steps_per_second": 23.532
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 76700,
+  "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": 2.340951304214108e+17,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}