End of training

README.md

@@ -1,13 +1,28 @@
 ---
 library_name: transformers
+language:
+- en
 base_model: gokulsrinivasagan/bert_tiny_lda_100_v1
 tags:
 - generated_from_trainer
+datasets:
+- glue
 metrics:
 - accuracy
 model-index:
 - name: bert_tiny_lda_100_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: 0.7035191212367778
 ---
 
 <!-- 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. -->
 
 # bert_tiny_lda_100_v1_mnli
 
-This model is a fine-tuned version of [gokulsrinivasagan/bert_tiny_lda_100_v1](https://huggingface.co/gokulsrinivasagan/bert_tiny_lda_100_v1) on an unknown dataset.
+This model is a fine-tuned version of [gokulsrinivasagan/bert_tiny_lda_100_v1](https://huggingface.co/gokulsrinivasagan/bert_tiny_lda_100_v1) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8340
-- Accuracy: 0.7018
+- Loss: 0.6947
+- Accuracy: 0.7035
 
 ## Model description
 

all_results.json

@@ -1,22 +1,22 @@
 {
-    "epoch": 14.0,
-    "epoch_mm": 14.0,
-    "eval_accuracy": 0.3273560876209883,
-    "eval_accuracy_mm": 0.3295362082994304,
-    "eval_loss": 1.0962358713150024,
-    "eval_loss_mm": 1.0962363481521606,
-    "eval_runtime": 2.9161,
-    "eval_runtime_mm": 2.9202,
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.6960774325012735,
+    "eval_accuracy_mm": 0.7035191212367778,
+    "eval_loss": 0.7240264415740967,
+    "eval_loss_mm": 0.6947348117828369,
+    "eval_runtime": 2.935,
+    "eval_runtime_mm": 2.9443,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 3365.834,
-    "eval_samples_per_second_mm": 3366.857,
-    "eval_steps_per_second": 13.374,
-    "eval_steps_per_second_mm": 13.355,
-    "total_flos": 1.4417705411687424e+17,
-    "train_loss": 1.0986616817082877,
-    "train_runtime": 2888.6025,
+    "eval_samples_per_second": 3344.081,
+    "eval_samples_per_second_mm": 3339.332,
+    "eval_steps_per_second": 13.288,
+    "eval_steps_per_second_mm": 13.246,
+    "total_flos": 1.029836100834816e+17,
+    "train_loss": 0.6466208924215185,
+    "train_runtime": 2106.5383,
     "train_samples": 392702,
-    "train_samples_per_second": 6797.439,
-    "train_steps_per_second": 26.553
+    "train_samples_per_second": 9321.027,
+    "train_steps_per_second": 36.41
 }

eval_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 14.0,
-    "epoch_mm": 14.0,
-    "eval_accuracy": 0.3273560876209883,
-    "eval_accuracy_mm": 0.3295362082994304,
-    "eval_loss": 1.0962358713150024,
-    "eval_loss_mm": 1.0962363481521606,
-    "eval_runtime": 2.9161,
-    "eval_runtime_mm": 2.9202,
+    "epoch": 10.0,
+    "epoch_mm": 10.0,
+    "eval_accuracy": 0.6960774325012735,
+    "eval_accuracy_mm": 0.7035191212367778,
+    "eval_loss": 0.7240264415740967,
+    "eval_loss_mm": 0.6947348117828369,
+    "eval_runtime": 2.935,
+    "eval_runtime_mm": 2.9443,
     "eval_samples": 9815,
     "eval_samples_mm": 9832,
-    "eval_samples_per_second": 3365.834,
-    "eval_samples_per_second_mm": 3366.857,
-    "eval_steps_per_second": 13.374,
-    "eval_steps_per_second_mm": 13.355
+    "eval_samples_per_second": 3344.081,
+    "eval_samples_per_second_mm": 3339.332,
+    "eval_steps_per_second": 13.288,
+    "eval_steps_per_second_mm": 13.246
 }

logs/events.out.tfevents.1733329564.ki-g0008.1208741.35

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa90e774d90cb23cca1ae769ee9eb61c52e3878e641a87873be0adacd03dbf64
+size 734

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 14.0,
-    "total_flos": 1.4417705411687424e+17,
-    "train_loss": 1.0986616817082877,
-    "train_runtime": 2888.6025,
+    "epoch": 10.0,
+    "total_flos": 1.029836100834816e+17,
+    "train_loss": 0.6466208924215185,
+    "train_runtime": 2106.5383,
     "train_samples": 392702,
-    "train_samples_per_second": 6797.439,
-    "train_steps_per_second": 26.553
+    "train_samples_per_second": 9321.027,
+    "train_steps_per_second": 36.41
 }

trainer_state.json

@@ -1,245 +1,181 @@
 {
-  "best_metric": 1.0962358713150024,
-  "best_model_checkpoint": "bert_tiny_lda_100_v1_mnli/checkpoint-13806",
-  "epoch": 14.0,
+  "best_metric": 0.7240264415740967,
+  "best_model_checkpoint": "bert_tiny_lda_100_v1_mnli/checkpoint-7670",
+  "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 21476,
+  "global_step": 15340,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 0.020302889868617058,
-      "learning_rate": 0.00098,
-      "loss": 1.0996,
+      "grad_norm": 1.503577470779419,
+      "learning_rate": 4.9e-05,
+      "loss": 0.9642,
       "step": 1534
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0992883443832397,
-      "eval_runtime": 2.9221,
-      "eval_samples_per_second": 3358.897,
-      "eval_steps_per_second": 13.347,
+      "eval_accuracy": 0.6113092205807438,
+      "eval_loss": 0.8560646176338196,
+      "eval_runtime": 3.0082,
+      "eval_samples_per_second": 3262.723,
+      "eval_steps_per_second": 12.964,
       "step": 1534
     },
     {
       "epoch": 2.0,
-      "grad_norm": 0.11855403333902359,
-      "learning_rate": 0.00096,
-      "loss": 1.0986,
+      "grad_norm": 1.652482032775879,
+      "learning_rate": 4.8e-05,
+      "loss": 0.8313,
       "step": 3068
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.1015625,
-      "eval_runtime": 2.9167,
-      "eval_samples_per_second": 3365.108,
-      "eval_steps_per_second": 13.371,
+      "eval_accuracy": 0.6519612837493632,
+      "eval_loss": 0.781214714050293,
+      "eval_runtime": 2.9745,
+      "eval_samples_per_second": 3299.719,
+      "eval_steps_per_second": 13.111,
       "step": 3068
     },
     {
       "epoch": 3.0,
-      "grad_norm": 0.038195185363292694,
-      "learning_rate": 0.00094,
-      "loss": 1.0987,
+      "grad_norm": 1.7888683080673218,
+      "learning_rate": 4.7e-05,
+      "loss": 0.757,
       "step": 4602
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0989333391189575,
-      "eval_runtime": 2.8846,
-      "eval_samples_per_second": 3402.535,
-      "eval_steps_per_second": 13.52,
+      "eval_accuracy": 0.6742740703005604,
+      "eval_loss": 0.7461988925933838,
+      "eval_runtime": 3.0332,
+      "eval_samples_per_second": 3235.842,
+      "eval_steps_per_second": 12.858,
       "step": 4602
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.062206000089645386,
-      "learning_rate": 0.00092,
-      "loss": 1.0984,
+      "grad_norm": 1.9277328252792358,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.6957,
       "step": 6136
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.10163414478302,
-      "eval_runtime": 2.9197,
-      "eval_samples_per_second": 3361.698,
-      "eval_steps_per_second": 13.358,
+      "eval_accuracy": 0.6888436067244015,
+      "eval_loss": 0.7299773097038269,
+      "eval_runtime": 2.9878,
+      "eval_samples_per_second": 3285.078,
+      "eval_steps_per_second": 13.053,
       "step": 6136
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.07343439757823944,
-      "learning_rate": 0.0009000000000000001,
-      "loss": 1.0985,
+      "grad_norm": 1.847220540046692,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6441,
       "step": 7670
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0988649129867554,
-      "eval_runtime": 2.9524,
-      "eval_samples_per_second": 3324.402,
-      "eval_steps_per_second": 13.21,
+      "eval_accuracy": 0.6960774325012735,
+      "eval_loss": 0.7240264415740967,
+      "eval_runtime": 2.9211,
+      "eval_samples_per_second": 3360.083,
+      "eval_steps_per_second": 13.351,
       "step": 7670
     },
     {
       "epoch": 6.0,
-      "grad_norm": 0.047856856137514114,
-      "learning_rate": 0.00088,
-      "loss": 1.0988,
+      "grad_norm": 1.830507516860962,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.5971,
       "step": 9204
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.3544574630667346,
-      "eval_loss": 1.0988649129867554,
-      "eval_runtime": 2.9016,
-      "eval_samples_per_second": 3382.572,
-      "eval_steps_per_second": 13.441,
+      "eval_accuracy": 0.6970962812022414,
+      "eval_loss": 0.7375145554542542,
+      "eval_runtime": 2.9239,
+      "eval_samples_per_second": 3356.817,
+      "eval_steps_per_second": 13.338,
       "step": 9204
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.0791020318865776,
-      "learning_rate": 0.00086,
-      "loss": 1.0985,
+      "grad_norm": 2.2411386966705322,
+      "learning_rate": 4.3e-05,
+      "loss": 0.5545,
       "step": 10738
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.0968025922775269,
-      "eval_runtime": 2.8821,
-      "eval_samples_per_second": 3405.537,
-      "eval_steps_per_second": 13.532,
+      "eval_accuracy": 0.70412633723892,
+      "eval_loss": 0.7476160526275635,
+      "eval_runtime": 2.9231,
+      "eval_samples_per_second": 3357.702,
+      "eval_steps_per_second": 13.342,
       "step": 10738
     },
     {
       "epoch": 8.0,
-      "grad_norm": 0.016667626798152924,
-      "learning_rate": 0.00084,
-      "loss": 1.0985,
+      "grad_norm": 2.366934299468994,
+      "learning_rate": 4.2e-05,
+      "loss": 0.5132,
       "step": 12272
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.0990766286849976,
-      "eval_runtime": 2.9132,
-      "eval_samples_per_second": 3369.162,
-      "eval_steps_per_second": 13.387,
+      "eval_accuracy": 0.7019867549668874,
+      "eval_loss": 0.7631875872612,
+      "eval_runtime": 2.9491,
+      "eval_samples_per_second": 3328.154,
+      "eval_steps_per_second": 13.224,
       "step": 12272
     },
     {
       "epoch": 9.0,
-      "grad_norm": 0.006579674780368805,
-      "learning_rate": 0.00082,
-      "loss": 1.0988,
+      "grad_norm": 2.796860456466675,
+      "learning_rate": 4.1e-05,
+      "loss": 0.4736,
       "step": 13806
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0962358713150024,
-      "eval_runtime": 2.9177,
-      "eval_samples_per_second": 3363.934,
-      "eval_steps_per_second": 13.367,
+      "eval_accuracy": 0.7009679062659195,
+      "eval_loss": 0.8263298869132996,
+      "eval_runtime": 2.9219,
+      "eval_samples_per_second": 3359.087,
+      "eval_steps_per_second": 13.347,
       "step": 13806
     },
     {
       "epoch": 10.0,
-      "grad_norm": 0.045499037951231,
-      "learning_rate": 0.0008,
-      "loss": 1.0986,
+      "grad_norm": 3.2133843898773193,
+      "learning_rate": 4e-05,
+      "loss": 0.4355,
       "step": 15340
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0992168188095093,
-      "eval_runtime": 2.9033,
-      "eval_samples_per_second": 3380.602,
-      "eval_steps_per_second": 13.433,
+      "eval_accuracy": 0.7017829852266938,
+      "eval_loss": 0.8340126276016235,
+      "eval_runtime": 2.9325,
+      "eval_samples_per_second": 3347.006,
+      "eval_steps_per_second": 13.299,
       "step": 15340
     },
     {
-      "epoch": 11.0,
-      "grad_norm": 0.009605828672647476,
-      "learning_rate": 0.0007800000000000001,
-      "loss": 1.0986,
-      "step": 16874
-    },
-    {
-      "epoch": 11.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0990049839019775,
-      "eval_runtime": 2.9296,
-      "eval_samples_per_second": 3350.335,
-      "eval_steps_per_second": 13.313,
-      "step": 16874
-    },
-    {
-      "epoch": 12.0,
-      "grad_norm": 0.03646823763847351,
-      "learning_rate": 0.00076,
-      "loss": 1.0984,
-      "step": 18408
-    },
-    {
-      "epoch": 12.0,
-      "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.0990766286849976,
-      "eval_runtime": 2.9172,
-      "eval_samples_per_second": 3364.556,
-      "eval_steps_per_second": 13.369,
-      "step": 18408
-    },
-    {
-      "epoch": 13.0,
-      "grad_norm": 0.09090559184551239,
-      "learning_rate": 0.00074,
-      "loss": 1.0986,
-      "step": 19942
-    },
-    {
-      "epoch": 13.0,
-      "eval_accuracy": 0.31818644931227713,
-      "eval_loss": 1.0965192317962646,
-      "eval_runtime": 2.9196,
-      "eval_samples_per_second": 3361.754,
-      "eval_steps_per_second": 13.358,
-      "step": 19942
-    },
-    {
-      "epoch": 14.0,
-      "grad_norm": 0.03515646234154701,
-      "learning_rate": 0.0007199999999999999,
-      "loss": 1.0987,
-      "step": 21476
-    },
-    {
-      "epoch": 14.0,
-      "eval_accuracy": 0.3273560876209883,
-      "eval_loss": 1.0987216234207153,
-      "eval_runtime": 2.9407,
-      "eval_samples_per_second": 3337.663,
-      "eval_steps_per_second": 13.262,
-      "step": 21476
-    },
-    {
-      "epoch": 14.0,
-      "step": 21476,
-      "total_flos": 1.4417705411687424e+17,
-      "train_loss": 1.0986616817082877,
-      "train_runtime": 2888.6025,
-      "train_samples_per_second": 6797.439,
-      "train_steps_per_second": 26.553
+      "epoch": 10.0,
+      "step": 15340,
+      "total_flos": 1.029836100834816e+17,
+      "train_loss": 0.6466208924215185,
+      "train_runtime": 2106.5383,
+      "train_samples_per_second": 9321.027,
+      "train_steps_per_second": 36.41
     }
   ],
   "logging_steps": 1,
@@ -268,7 +204,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.4417705411687424e+17,
+  "total_flos": 1.029836100834816e+17,
   "train_batch_size": 256,
   "trial_name": null,
   "trial_params": null