End of training

README.md

@@ -1,13 +1,28 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/tiny_bert_km_5_v2
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: tiny_bert_km_5_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: 0.6934499593165175
 ---
 
 <!-- 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. -->
 
 # tiny_bert_km_5_v2_mnli
 
-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 MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8758
-- Accuracy: 0.6904
+- Loss: 0.7252
+- Accuracy: 0.6934
 
 ## Model description
 

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 12.0,
+    "epoch_mm": 12.0,
+    "eval_accuracy": 0.6919001528273051,
+    "eval_accuracy_mm": 0.6934499593165175,
+    "eval_loss": 0.7364546060562134,
+    "eval_loss_mm": 0.7251723408699036,
+    "eval_runtime": 2.8329,
+    "eval_runtime_mm": 2.8569,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 3464.698,
+    "eval_samples_per_second_mm": 3441.513,
+    "eval_steps_per_second": 13.767,
+    "eval_steps_per_second_mm": 13.651,
+    "total_flos": 1.2358033210017792e+17,
+    "train_loss": 0.6455690631758695,
+    "train_runtime": 2475.6151,
+    "train_samples": 392702,
+    "train_samples_per_second": 7931.403,
+    "train_steps_per_second": 30.982
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 12.0,
+    "epoch_mm": 12.0,
+    "eval_accuracy": 0.6919001528273051,
+    "eval_accuracy_mm": 0.6934499593165175,
+    "eval_loss": 0.7364546060562134,
+    "eval_loss_mm": 0.7251723408699036,
+    "eval_runtime": 2.8329,
+    "eval_runtime_mm": 2.8569,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 3464.698,
+    "eval_samples_per_second_mm": 3441.513,
+    "eval_steps_per_second": 13.767,
+    "eval_steps_per_second_mm": 13.651
+}

logs/events.out.tfevents.1745273387.s_005_m.2836625.17

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:446be10d49a80b8323c6e9668529b353bfb6e3b3deba0a75f48bf4b2ef2b6422
+size 698

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 12.0,
+    "total_flos": 1.2358033210017792e+17,
+    "train_loss": 0.6455690631758695,
+    "train_runtime": 2475.6151,
+    "train_samples": 392702,
+    "train_samples_per_second": 7931.403,
+    "train_steps_per_second": 30.982
+}

trainer_state.json

@@ -0,0 +1,244 @@
+{
+  "best_global_step": 10738,
+  "best_metric": 0.7364546060562134,
+  "best_model_checkpoint": "tiny_bert_km_5_v2_mnli/checkpoint-10738",
+  "epoch": 12.0,
+  "eval_steps": 500,
+  "global_step": 18408,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4700506925582886,
+      "learning_rate": 4.9e-05,
+      "loss": 0.9851,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5840040753948039,
+      "eval_loss": 0.8961641788482666,
+      "eval_runtime": 2.8662,
+      "eval_samples_per_second": 3424.409,
+      "eval_steps_per_second": 13.607,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.7692183256149292,
+      "learning_rate": 4.8e-05,
+      "loss": 0.8564,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.642791645440652,
+      "eval_loss": 0.8020933270454407,
+      "eval_runtime": 2.7975,
+      "eval_samples_per_second": 3508.465,
+      "eval_steps_per_second": 13.941,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.8865474462509155,
+      "learning_rate": 4.7e-05,
+      "loss": 0.7776,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.6653082017320427,
+      "eval_loss": 0.7620525360107422,
+      "eval_runtime": 2.8014,
+      "eval_samples_per_second": 3503.618,
+      "eval_steps_per_second": 13.922,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.254871129989624,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.7255,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6747834946510444,
+      "eval_loss": 0.7468968629837036,
+      "eval_runtime": 2.778,
+      "eval_samples_per_second": 3533.176,
+      "eval_steps_per_second": 14.039,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 2.165499448776245,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6815,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6851757514009169,
+      "eval_loss": 0.7374223470687866,
+      "eval_runtime": 2.782,
+      "eval_samples_per_second": 3528.034,
+      "eval_steps_per_second": 14.019,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.329120397567749,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.6403,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6903718797758532,
+      "eval_loss": 0.7439625263214111,
+      "eval_runtime": 2.7994,
+      "eval_samples_per_second": 3506.073,
+      "eval_steps_per_second": 13.931,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.04762601852417,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6029,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6919001528273051,
+      "eval_loss": 0.7364546060562134,
+      "eval_runtime": 2.8157,
+      "eval_samples_per_second": 3485.755,
+      "eval_steps_per_second": 13.851,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.1585097312927246,
+      "learning_rate": 4.2e-05,
+      "loss": 0.5657,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6908813041263372,
+      "eval_loss": 0.7683497667312622,
+      "eval_runtime": 2.7729,
+      "eval_samples_per_second": 3539.657,
+      "eval_steps_per_second": 14.065,
+      "step": 12272
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 2.760887861251831,
+      "learning_rate": 4.1e-05,
+      "loss": 0.5311,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.6909831889964341,
+      "eval_loss": 0.8052825331687927,
+      "eval_runtime": 2.8244,
+      "eval_samples_per_second": 3475.111,
+      "eval_steps_per_second": 13.808,
+      "step": 13806
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 2.903724431991577,
+      "learning_rate": 4e-05,
+      "loss": 0.4934,
+      "step": 15340
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.6928171166581762,
+      "eval_loss": 0.8145983815193176,
+      "eval_runtime": 2.8224,
+      "eval_samples_per_second": 3477.505,
+      "eval_steps_per_second": 13.818,
+      "step": 15340
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 3.648442268371582,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.4595,
+      "step": 16874
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.6911869587366276,
+      "eval_loss": 0.8572707176208496,
+      "eval_runtime": 2.8013,
+      "eval_samples_per_second": 3503.67,
+      "eval_steps_per_second": 13.922,
+      "step": 16874
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 3.2090775966644287,
+      "learning_rate": 3.8e-05,
+      "loss": 0.4278,
+      "step": 18408
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.6903718797758532,
+      "eval_loss": 0.8758233189582825,
+      "eval_runtime": 2.8088,
+      "eval_samples_per_second": 3494.436,
+      "eval_steps_per_second": 13.885,
+      "step": 18408
+    },
+    {
+      "epoch": 12.0,
+      "step": 18408,
+      "total_flos": 1.2358033210017792e+17,
+      "train_loss": 0.6455690631758695,
+      "train_runtime": 2475.6151,
+      "train_samples_per_second": 7931.403,
+      "train_steps_per_second": 30.982
+    }
+  ],
+  "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": 1.2358033210017792e+17,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}