End of training

README.md

@@ -14,7 +14,7 @@ model-index:
       name: Masked Language Modeling
       type: fill-mask
     dataset:
-      name: wikitext
+      name: wikitext wikitext-2-raw-v1
       type: wikitext
       config: wikitext-2-raw-v1
       split: validation
@@ -22,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.7288328898061153
+      value: 0.7255275697753574
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -30,10 +30,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # mlm
 
-This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the wikitext dataset.
+This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the wikitext wikitext-2-raw-v1 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.2628
-- Accuracy: 0.7288
+- Loss: 1.2799
+- Accuracy: 0.7255
 
 ## Model description
 

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.7255275697753574,
+    "eval_loss": 1.2798649072647095,
+    "eval_runtime": 13.3316,
+    "eval_samples": 496,
+    "eval_samples_per_second": 37.205,
+    "eval_steps_per_second": 4.651,
+    "perplexity": 3.596153878488844,
+    "total_flos": 3789443078682624.0,
+    "train_loss": 1.420832945505778,
+    "train_runtime": 1064.4338,
+    "train_samples": 4798,
+    "train_samples_per_second": 13.523,
+    "train_steps_per_second": 0.423
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.7255275697753574,
+    "eval_loss": 1.2798649072647095,
+    "eval_runtime": 13.3316,
+    "eval_samples": 496,
+    "eval_samples_per_second": 37.205,
+    "eval_steps_per_second": 4.651,
+    "perplexity": 3.596153878488844
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "total_flos": 3789443078682624.0,
+    "train_loss": 1.420832945505778,
+    "train_runtime": 1064.4338,
+    "train_samples": 4798,
+    "train_samples_per_second": 13.523,
+    "train_steps_per_second": 0.423
+}

trainer_state.json

@@ -0,0 +1,325 @@
+{
+  "best_metric": 0.7289605637620273,
+  "best_model_checkpoint": "D:/1_SyscoPY_D/NLP/Data/Transformers_Hug/checkpoint/roberta-base\\checkpoint-150",
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 450,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.8568,
+      "step": 10
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.7794,
+      "step": 20
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 1.5937,
+      "step": 30
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.557,
+      "step": 40
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 1.5075,
+      "step": 50
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 1.4945,
+      "step": 60
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8148148148148147e-05,
+      "loss": 1.4764,
+      "step": 70
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7481481481481482e-05,
+      "loss": 1.4568,
+      "step": 80
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.6740740740740743e-05,
+      "loss": 1.4276,
+      "step": 90
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.4126,
+      "step": 100
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.525925925925926e-05,
+      "loss": 1.3712,
+      "step": 110
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.451851851851852e-05,
+      "loss": 1.4503,
+      "step": 120
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.377777777777778e-05,
+      "loss": 1.4199,
+      "step": 130
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.303703703703704e-05,
+      "loss": 1.469,
+      "step": 140
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2296296296296297e-05,
+      "loss": 1.3784,
+      "step": 150
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7289605637620273,
+      "eval_loss": 1.282206416130066,
+      "eval_runtime": 13.5027,
+      "eval_samples_per_second": 36.733,
+      "eval_steps_per_second": 4.592,
+      "step": 150
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.155555555555556e-05,
+      "loss": 1.3744,
+      "step": 160
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0814814814814817e-05,
+      "loss": 1.4279,
+      "step": 170
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.014814814814815e-05,
+      "loss": 1.4016,
+      "step": 180
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9407407407407407e-05,
+      "loss": 1.3812,
+      "step": 190
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.415,
+      "step": 200
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.7925925925925927e-05,
+      "loss": 1.3856,
+      "step": 210
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.7185185185185185e-05,
+      "loss": 1.3874,
+      "step": 220
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.6444444444444444e-05,
+      "loss": 1.3965,
+      "step": 230
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.5703703703703705e-05,
+      "loss": 1.4211,
+      "step": 240
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.4962962962962964e-05,
+      "loss": 1.3866,
+      "step": 250
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 1.3759,
+      "step": 260
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.3481481481481482e-05,
+      "loss": 1.4049,
+      "step": 270
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2740740740740742e-05,
+      "loss": 1.3919,
+      "step": 280
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2e-05,
+      "loss": 1.3781,
+      "step": 290
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.125925925925926e-05,
+      "loss": 1.3804,
+      "step": 300
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7273364801078894,
+      "eval_loss": 1.275496482849121,
+      "eval_runtime": 13.2216,
+      "eval_samples_per_second": 37.514,
+      "eval_steps_per_second": 4.689,
+      "step": 300
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.051851851851852e-05,
+      "loss": 1.3812,
+      "step": 310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 1.4056,
+      "step": 320
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.037037037037039e-06,
+      "loss": 1.3889,
+      "step": 330
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.296296296296295e-06,
+      "loss": 1.3811,
+      "step": 340
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 7.555555555555555e-06,
+      "loss": 1.3451,
+      "step": 350
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 6.814814814814815e-06,
+      "loss": 1.3535,
+      "step": 360
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 6.0740740740740745e-06,
+      "loss": 1.3702,
+      "step": 370
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.3499,
+      "step": 380
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 4.592592592592593e-06,
+      "loss": 1.3318,
+      "step": 390
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.851851851851852e-06,
+      "loss": 1.3391,
+      "step": 400
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.111111111111111e-06,
+      "loss": 1.3648,
+      "step": 410
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3703703703703703e-06,
+      "loss": 1.3147,
+      "step": 420
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.6296296296296297e-06,
+      "loss": 1.3406,
+      "step": 430
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 1.3526,
+      "step": 440
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.4814814814814815e-07,
+      "loss": 1.3586,
+      "step": 450
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7288328898061153,
+      "eval_loss": 1.2627531290054321,
+      "eval_runtime": 13.2416,
+      "eval_samples_per_second": 37.458,
+      "eval_steps_per_second": 4.682,
+      "step": 450
+    },
+    {
+      "epoch": 3.0,
+      "step": 450,
+      "total_flos": 3789443078682624.0,
+      "train_loss": 1.420832945505778,
+      "train_runtime": 1064.4338,
+      "train_samples_per_second": 13.523,
+      "train_steps_per_second": 0.423
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 450,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "total_flos": 3789443078682624.0,
+  "trial_name": null,
+  "trial_params": null
+}