End of training

README.md

@@ -1,12 +1,27 @@
 ---
+language:
+- en
 base_model: Hartunka/tiny_bert_km_100_v1
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: tiny_bert_km_100_v1_rte
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE RTE
+      type: glue
+      args: rte
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.5270758122743683
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -14,10 +29,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tiny_bert_km_100_v1_rte
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_km_100_v1](https://huggingface.co/Hartunka/tiny_bert_km_100_v1) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_km_100_v1](https://huggingface.co/Hartunka/tiny_bert_km_100_v1) on the GLUE RTE dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8417
-- Accuracy: 0.5307
+- Loss: 0.6949
+- Accuracy: 0.5271
 
 ## Model description
 

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5270758122743683,
+    "eval_loss": 0.694903552532196,
+    "eval_runtime": 0.141,
+    "eval_samples": 277,
+    "eval_samples_per_second": 1965.206,
+    "eval_steps_per_second": 14.189,
+    "total_flos": 522373675991040.0,
+    "train_loss": 0.6144436895847321,
+    "train_runtime": 21.1487,
+    "train_samples": 2490,
+    "train_samples_per_second": 5886.889,
+    "train_steps_per_second": 23.642
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "eval_accuracy": 0.5270758122743683,
+    "eval_loss": 0.694903552532196,
+    "eval_runtime": 0.141,
+    "eval_samples": 277,
+    "eval_samples_per_second": 1965.206,
+    "eval_steps_per_second": 14.189
+}

logs/events.out.tfevents.1744820845.s_004_m

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d69d49a46b3b2ec6f3fc90de98623c9a4aa377a8b898877d9f31bae0f203d86
+size 249

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "total_flos": 522373675991040.0,
+    "train_loss": 0.6144436895847321,
+    "train_runtime": 21.1487,
+    "train_samples": 2490,
+    "train_samples_per_second": 5886.889,
+    "train_steps_per_second": 23.642
+}

trainer_state.json

@@ -0,0 +1,158 @@
+{
+  "best_metric": 0.694903552532196,
+  "best_model_checkpoint": "tiny_bert_km_100_v1_rte/checkpoint-30",
+  "epoch": 8.0,
+  "eval_steps": 500,
+  "global_step": 80,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6791782975196838,
+      "learning_rate": 4.9e-05,
+      "loss": 0.699,
+      "step": 10
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5270758122743683,
+      "eval_loss": 0.6959471106529236,
+      "eval_runtime": 0.1358,
+      "eval_samples_per_second": 2040.346,
+      "eval_steps_per_second": 14.732,
+      "step": 10
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7037239074707031,
+      "learning_rate": 4.8e-05,
+      "loss": 0.6852,
+      "step": 20
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.4981949458483754,
+      "eval_loss": 0.7013058662414551,
+      "eval_runtime": 0.1226,
+      "eval_samples_per_second": 2259.809,
+      "eval_steps_per_second": 16.316,
+      "step": 20
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6154415011405945,
+      "learning_rate": 4.7e-05,
+      "loss": 0.6713,
+      "step": 30
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5270758122743683,
+      "eval_loss": 0.694903552532196,
+      "eval_runtime": 0.1148,
+      "eval_samples_per_second": 2412.236,
+      "eval_steps_per_second": 17.417,
+      "step": 30
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.636819064617157,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.6531,
+      "step": 40
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5451263537906137,
+      "eval_loss": 0.7031038403511047,
+      "eval_runtime": 0.1179,
+      "eval_samples_per_second": 2348.643,
+      "eval_steps_per_second": 16.958,
+      "step": 40
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.8013365268707275,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6265,
+      "step": 50
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.5234657039711191,
+      "eval_loss": 0.708032488822937,
+      "eval_runtime": 0.1174,
+      "eval_samples_per_second": 2359.336,
+      "eval_steps_per_second": 17.035,
+      "step": 50
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.9887241125106812,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.5887,
+      "step": 60
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.516245487364621,
+      "eval_loss": 0.7479658126831055,
+      "eval_runtime": 0.1163,
+      "eval_samples_per_second": 2380.993,
+      "eval_steps_per_second": 17.191,
+      "step": 60
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.28850519657135,
+      "learning_rate": 4.3e-05,
+      "loss": 0.5347,
+      "step": 70
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.5342960288808665,
+      "eval_loss": 0.7832542657852173,
+      "eval_runtime": 0.125,
+      "eval_samples_per_second": 2216.571,
+      "eval_steps_per_second": 16.004,
+      "step": 70
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.8739584684371948,
+      "learning_rate": 4.2e-05,
+      "loss": 0.457,
+      "step": 80
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.5306859205776173,
+      "eval_loss": 0.8416646718978882,
+      "eval_runtime": 0.1169,
+      "eval_samples_per_second": 2369.547,
+      "eval_steps_per_second": 17.109,
+      "step": 80
+    },
+    {
+      "epoch": 8.0,
+      "step": 80,
+      "total_flos": 522373675991040.0,
+      "train_loss": 0.6144436895847321,
+      "train_runtime": 21.1487,
+      "train_samples_per_second": 5886.889,
+      "train_steps_per_second": 23.642
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 500,
+  "total_flos": 522373675991040.0,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}