End of training

README.md

@@ -1,13 +1,28 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: Hartunka/tiny_bert_rand_20_v1
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: tiny_bert_rand_20_v1_mnli
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE MNLI
+      type: glue
+      args: mnli
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.5988608624898292
 ---
 
 <!-- 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_rand_20_v1_mnli
 
-This model is a fine-tuned version of [Hartunka/tiny_bert_rand_20_v1](https://huggingface.co/Hartunka/tiny_bert_rand_20_v1) on an unknown dataset.
+This model is a fine-tuned version of [Hartunka/tiny_bert_rand_20_v1](https://huggingface.co/Hartunka/tiny_bert_rand_20_v1) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.1121
-- Accuracy: 0.6019
+- Loss: 0.8709
+- Accuracy: 0.5989
 
 ## Model description
 

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 9.0,
+    "epoch_mm": 9.0,
+    "eval_accuracy": 0.6011207335710647,
+    "eval_accuracy_mm": 0.5988608624898292,
+    "eval_loss": 0.8717783093452454,
+    "eval_loss_mm": 0.8709185123443604,
+    "eval_runtime": 2.8752,
+    "eval_runtime_mm": 2.9609,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 3413.661,
+    "eval_samples_per_second_mm": 3320.563,
+    "eval_steps_per_second": 13.564,
+    "eval_steps_per_second_mm": 13.171,
+    "total_flos": 9.268524907513344e+16,
+    "train_loss": 0.7542917184444263,
+    "train_runtime": 1890.837,
+    "train_samples": 392702,
+    "train_samples_per_second": 10384.343,
+    "train_steps_per_second": 40.564
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 9.0,
+    "epoch_mm": 9.0,
+    "eval_accuracy": 0.6011207335710647,
+    "eval_accuracy_mm": 0.5988608624898292,
+    "eval_loss": 0.8717783093452454,
+    "eval_loss_mm": 0.8709185123443604,
+    "eval_runtime": 2.8752,
+    "eval_runtime_mm": 2.9609,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 3413.661,
+    "eval_samples_per_second_mm": 3320.563,
+    "eval_steps_per_second": 13.564,
+    "eval_steps_per_second_mm": 13.171
+}

logs/events.out.tfevents.1744656779.s_005_m.2701049.107

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6689072a084c0c6285c338463fc52a96d8b54e01975bafd7d519399f16beb12e
+size 686

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 9.0,
+    "total_flos": 9.268524907513344e+16,
+    "train_loss": 0.7542917184444263,
+    "train_runtime": 1890.837,
+    "train_samples": 392702,
+    "train_samples_per_second": 10384.343,
+    "train_steps_per_second": 40.564
+}

trainer_state.json

@@ -0,0 +1,196 @@
+{
+  "best_global_step": 6136,
+  "best_metric": 0.8717783093452454,
+  "best_model_checkpoint": "tiny_bert_rand_20_v1_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": 1.116929531097412,
+      "learning_rate": 4.9e-05,
+      "loss": 0.9987,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5349974528782476,
+      "eval_loss": 0.9482256770133972,
+      "eval_runtime": 3.0048,
+      "eval_samples_per_second": 3266.443,
+      "eval_steps_per_second": 12.979,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5099765062332153,
+      "learning_rate": 4.8e-05,
+      "loss": 0.9153,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5830871115639328,
+      "eval_loss": 0.8893212676048279,
+      "eval_runtime": 2.8696,
+      "eval_samples_per_second": 3420.282,
+      "eval_steps_per_second": 13.591,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.534440279006958,
+      "learning_rate": 4.7e-05,
+      "loss": 0.8571,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5966377992868059,
+      "eval_loss": 0.8751558065414429,
+      "eval_runtime": 2.8585,
+      "eval_samples_per_second": 3433.572,
+      "eval_steps_per_second": 13.643,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.8849776983261108,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.8033,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.6011207335710647,
+      "eval_loss": 0.8717783093452454,
+      "eval_runtime": 2.9065,
+      "eval_samples_per_second": 3376.859,
+      "eval_steps_per_second": 13.418,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.8985122442245483,
+      "learning_rate": 4.5e-05,
+      "loss": 0.7482,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6078451349974529,
+      "eval_loss": 0.8746739625930786,
+      "eval_runtime": 2.9141,
+      "eval_samples_per_second": 3368.133,
+      "eval_steps_per_second": 13.383,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 2.6544694900512695,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.6947,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6129393785022924,
+      "eval_loss": 0.9125290513038635,
+      "eval_runtime": 2.8702,
+      "eval_samples_per_second": 3419.607,
+      "eval_steps_per_second": 13.588,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.5749664306640625,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6416,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6154865002547122,
+      "eval_loss": 0.9405058026313782,
+      "eval_runtime": 2.8943,
+      "eval_samples_per_second": 3391.14,
+      "eval_steps_per_second": 13.475,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 2.8357646465301514,
+      "learning_rate": 4.2e-05,
+      "loss": 0.5894,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6038716250636781,
+      "eval_loss": 1.0252567529678345,
+      "eval_runtime": 2.8812,
+      "eval_samples_per_second": 3406.525,
+      "eval_steps_per_second": 13.536,
+      "step": 12272
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 3.2707016468048096,
+      "learning_rate": 4.1e-05,
+      "loss": 0.5403,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.601935812531839,
+      "eval_loss": 1.1121231317520142,
+      "eval_runtime": 2.903,
+      "eval_samples_per_second": 3380.968,
+      "eval_steps_per_second": 13.434,
+      "step": 13806
+    },
+    {
+      "epoch": 9.0,
+      "step": 13806,
+      "total_flos": 9.268524907513344e+16,
+      "train_loss": 0.7542917184444263,
+      "train_runtime": 1890.837,
+      "train_samples_per_second": 10384.343,
+      "train_steps_per_second": 40.564
+    }
+  ],
+  "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": 9.268524907513344e+16,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}