End of training

README.md

@@ -1,14 +1,29 @@
 ---
 library_name: transformers
+language:
+- en
 license: apache-2.0
 base_model: google/bert_uncased_L-4_H-512_A-8
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: bert_uncased_L-4_H-512_A-8_mnli
-  results: []
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE MNLI
+      type: glue
+      args: mnli
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.7811228641171685
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -16,10 +31,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # bert_uncased_L-4_H-512_A-8_mnli
 
-This model is a fine-tuned version of [google/bert_uncased_L-4_H-512_A-8](https://huggingface.co/google/bert_uncased_L-4_H-512_A-8) on an unknown dataset.
+This model is a fine-tuned version of [google/bert_uncased_L-4_H-512_A-8](https://huggingface.co/google/bert_uncased_L-4_H-512_A-8) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7132
-- Accuracy: 0.7763
+- Loss: 0.5537
+- Accuracy: 0.7811
 
 ## Model description
 

all_results.json

@@ -0,0 +1,22 @@
+{
+    "epoch": 8.0,
+    "epoch_mm": 8.0,
+    "eval_accuracy": 0.7793173713703515,
+    "eval_accuracy_mm": 0.7811228641171685,
+    "eval_loss": 0.5685115456581116,
+    "eval_loss_mm": 0.5536965727806091,
+    "eval_runtime": 3.9076,
+    "eval_runtime_mm": 3.8796,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2511.776,
+    "eval_samples_per_second_mm": 2534.313,
+    "eval_steps_per_second": 9.981,
+    "eval_steps_per_second_mm": 10.053,
+    "total_flos": 6.212741576306688e+16,
+    "train_loss": 0.43437760563220007,
+    "train_runtime": 1957.6876,
+    "train_samples": 392702,
+    "train_samples_per_second": 10029.741,
+    "train_steps_per_second": 39.179
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 8.0,
+    "epoch_mm": 8.0,
+    "eval_accuracy": 0.7793173713703515,
+    "eval_accuracy_mm": 0.7811228641171685,
+    "eval_loss": 0.5685115456581116,
+    "eval_loss_mm": 0.5536965727806091,
+    "eval_runtime": 3.9076,
+    "eval_runtime_mm": 3.8796,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 2511.776,
+    "eval_samples_per_second_mm": 2534.313,
+    "eval_steps_per_second": 9.981,
+    "eval_steps_per_second_mm": 10.053
+}

logs/events.out.tfevents.1733340411.ki-g0008.1761130.35

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66eb85a6c5ea6cdbc5b55a4725c6fc03482e6a89cea28188f06c4228726585b0
+size 734

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 8.0,
+    "total_flos": 6.212741576306688e+16,
+    "train_loss": 0.43437760563220007,
+    "train_runtime": 1957.6876,
+    "train_samples": 392702,
+    "train_samples_per_second": 10029.741,
+    "train_steps_per_second": 39.179
+}

trainer_state.json

@@ -0,0 +1,179 @@
+{
+  "best_metric": 0.5685115456581116,
+  "best_model_checkpoint": "bert_uncased_L-4_H-512_A-8_mnli/checkpoint-4602",
+  "epoch": 8.0,
+  "eval_steps": 500,
+  "global_step": 12272,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8079047203063965,
+      "learning_rate": 4.9e-05,
+      "loss": 0.7086,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7437595517065716,
+      "eval_loss": 0.6315475106239319,
+      "eval_runtime": 3.8782,
+      "eval_samples_per_second": 2530.843,
+      "eval_steps_per_second": 10.056,
+      "step": 1534
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.195190191268921,
+      "learning_rate": 4.8e-05,
+      "loss": 0.5704,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7651553744268976,
+      "eval_loss": 0.5782365798950195,
+      "eval_runtime": 3.8189,
+      "eval_samples_per_second": 2570.109,
+      "eval_steps_per_second": 10.212,
+      "step": 3068
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.5033040046691895,
+      "learning_rate": 4.7e-05,
+      "loss": 0.4967,
+      "step": 4602
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7793173713703515,
+      "eval_loss": 0.5685115456581116,
+      "eval_runtime": 4.1555,
+      "eval_samples_per_second": 2361.944,
+      "eval_steps_per_second": 9.385,
+      "step": 4602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.240811586380005,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.4356,
+      "step": 6136
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.7821701477330616,
+      "eval_loss": 0.5847141146659851,
+      "eval_runtime": 3.813,
+      "eval_samples_per_second": 2574.059,
+      "eval_steps_per_second": 10.228,
+      "step": 6136
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 4.033494472503662,
+      "learning_rate": 4.5e-05,
+      "loss": 0.3809,
+      "step": 7670
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.7855323484462557,
+      "eval_loss": 0.600440502166748,
+      "eval_runtime": 3.9079,
+      "eval_samples_per_second": 2511.562,
+      "eval_steps_per_second": 9.98,
+      "step": 7670
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 3.2427990436553955,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.3341,
+      "step": 9204
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.7836984207845135,
+      "eval_loss": 0.6375390887260437,
+      "eval_runtime": 3.8931,
+      "eval_samples_per_second": 2521.117,
+      "eval_steps_per_second": 10.018,
+      "step": 9204
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 2.801443576812744,
+      "learning_rate": 4.3e-05,
+      "loss": 0.2922,
+      "step": 10738
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7808456444218034,
+      "eval_loss": 0.6665855050086975,
+      "eval_runtime": 3.8712,
+      "eval_samples_per_second": 2535.361,
+      "eval_steps_per_second": 10.074,
+      "step": 10738
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 4.255834102630615,
+      "learning_rate": 4.2e-05,
+      "loss": 0.2565,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.7762608252674478,
+      "eval_loss": 0.7132487893104553,
+      "eval_runtime": 3.8681,
+      "eval_samples_per_second": 2537.445,
+      "eval_steps_per_second": 10.083,
+      "step": 12272
+    },
+    {
+      "epoch": 8.0,
+      "step": 12272,
+      "total_flos": 6.212741576306688e+16,
+      "train_loss": 0.43437760563220007,
+      "train_runtime": 1957.6876,
+      "train_samples_per_second": 10029.741,
+      "train_steps_per_second": 39.179
+    }
+  ],
+  "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": 6.212741576306688e+16,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}