End of training

README.md

@@ -3,6 +3,8 @@ license: mit
 base_model: roberta-base
 tags:
 - generated_from_trainer
+datasets:
+- fursov/gec_ner_val3
 metrics:
 - precision
 - recall
@@ -10,7 +12,26 @@ metrics:
 - accuracy
 model-index:
 - name: ner-gec-roberta-v3
-  results: []
+  results:
+  - task:
+      name: Token Classification
+      type: token-classification
+    dataset:
+      name: fursov/gec_ner_val3
+      type: fursov/gec_ner_val3
+    metrics:
+    - name: Precision
+      type: precision
+      value: 0.5705440070765149
+    - name: Recall
+      type: recall
+      value: 0.43481191856545776
+    - name: F1
+      type: f1
+      value: 0.493515436703776
+    - name: Accuracy
+      type: accuracy
+      value: 0.9566099116988466
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -18,13 +39,13 @@ should probably proofread and complete it, then remove this comment. -->
 
 # ner-gec-roberta-v3
 
-This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on an unknown dataset.
+This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the fursov/gec_ner_val3 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1783
-- Precision: 0.5595
-- Recall: 0.4328
-- F1: 0.4880
-- Accuracy: 0.9554
+- Loss: 0.1759
+- Precision: 0.5705
+- Recall: 0.4348
+- F1: 0.4935
+- Accuracy: 0.9566
 
 ## Model description
 

all_results.json

@@ -1,17 +1,17 @@
 {
-    "epoch": 5.0,
-    "eval_accuracy": 0.9483607249989437,
-    "eval_f1": 0.3637840032480715,
-    "eval_loss": 0.18947799503803253,
-    "eval_precision": 0.4573295222539812,
-    "eval_recall": 0.30200889847647294,
-    "eval_runtime": 7.5138,
+    "epoch": 10.0,
+    "eval_accuracy": 0.9566099116988466,
+    "eval_f1": 0.493515436703776,
+    "eval_loss": 0.1758583039045334,
+    "eval_precision": 0.5705440070765149,
+    "eval_recall": 0.43481191856545776,
+    "eval_runtime": 7.8131,
     "eval_samples": 4000,
-    "eval_samples_per_second": 532.352,
-    "eval_steps_per_second": 66.544,
-    "train_loss": 0.2075410250694521,
-    "train_runtime": 306.0993,
+    "eval_samples_per_second": 511.961,
+    "eval_steps_per_second": 63.995,
+    "train_loss": 0.06244980893376786,
+    "train_runtime": 304.7861,
     "train_samples": 55538,
-    "train_samples_per_second": 907.189,
-    "train_steps_per_second": 7.089
+    "train_samples_per_second": 1822.196,
+    "train_steps_per_second": 14.239
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 5.0,
-    "eval_accuracy": 0.9483607249989437,
-    "eval_f1": 0.3637840032480715,
-    "eval_loss": 0.18947799503803253,
-    "eval_precision": 0.4573295222539812,
-    "eval_recall": 0.30200889847647294,
-    "eval_runtime": 7.5138,
+    "epoch": 10.0,
+    "eval_accuracy": 0.9566099116988466,
+    "eval_f1": 0.493515436703776,
+    "eval_loss": 0.1758583039045334,
+    "eval_precision": 0.5705440070765149,
+    "eval_recall": 0.43481191856545776,
+    "eval_runtime": 7.8131,
     "eval_samples": 4000,
-    "eval_samples_per_second": 532.352,
-    "eval_steps_per_second": 66.544
+    "eval_samples_per_second": 511.961,
+    "eval_steps_per_second": 63.995
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 5.0,
-    "train_loss": 0.2075410250694521,
-    "train_runtime": 306.0993,
+    "epoch": 10.0,
+    "train_loss": 0.06244980893376786,
+    "train_runtime": 304.7861,
     "train_samples": 55538,
-    "train_samples_per_second": 907.189,
-    "train_steps_per_second": 7.089
+    "train_samples_per_second": 1822.196,
+    "train_steps_per_second": 14.239
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 5.0,
+  "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 2170,
+  "global_step": 4340,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -315,21 +315,327 @@
       "step": 2150
     },
     {
-      "epoch": 5.0,
-      "step": 2170,
-      "total_flos": 6755106697897248.0,
-      "train_loss": 0.2075410250694521,
-      "train_runtime": 306.0993,
-      "train_samples_per_second": 907.189,
-      "train_steps_per_second": 7.089
+      "epoch": 5.07,
+      "learning_rate": 2.4654377880184332e-05,
+      "loss": 0.1501,
+      "step": 2200
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.4078341013824887e-05,
+      "loss": 0.1533,
+      "step": 2250
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.350230414746544e-05,
+      "loss": 0.1555,
+      "step": 2300
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.2926267281105994e-05,
+      "loss": 0.1521,
+      "step": 2350
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.2350230414746546e-05,
+      "loss": 0.1549,
+      "step": 2400
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.1774193548387097e-05,
+      "loss": 0.1529,
+      "step": 2450
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 2.1198156682027652e-05,
+      "loss": 0.1546,
+      "step": 2500
+    },
+    {
+      "epoch": 5.76,
+      "eval_accuracy": 0.9504309434281127,
+      "eval_f1": 0.38409736308316433,
+      "eval_loss": 0.1856723576784134,
+      "eval_precision": 0.48227383863080686,
+      "eval_recall": 0.31913172441688015,
+      "eval_runtime": 7.5018,
+      "eval_samples_per_second": 533.207,
+      "eval_steps_per_second": 66.651,
+      "step": 2500
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.0622119815668204e-05,
+      "loss": 0.152,
+      "step": 2550
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.0046082949308755e-05,
+      "loss": 0.1527,
+      "step": 2600
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.9470046082949307e-05,
+      "loss": 0.1373,
+      "step": 2650
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1.8894009216589862e-05,
+      "loss": 0.1373,
+      "step": 2700
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.8317972350230417e-05,
+      "loss": 0.1355,
+      "step": 2750
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.1367,
+      "step": 2800
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.7165898617511524e-05,
+      "loss": 0.1344,
+      "step": 2850
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1.6589861751152075e-05,
+      "loss": 0.135,
+      "step": 2900
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.6013824884792627e-05,
+      "loss": 0.1348,
+      "step": 2950
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.543778801843318e-05,
+      "loss": 0.1343,
+      "step": 3000
+    },
+    {
+      "epoch": 6.91,
+      "eval_accuracy": 0.9534517723604715,
+      "eval_f1": 0.4423137378182962,
+      "eval_loss": 0.17842046916484833,
+      "eval_precision": 0.5302430751837196,
+      "eval_recall": 0.3793986787110692,
+      "eval_runtime": 7.7899,
+      "eval_samples_per_second": 513.483,
+      "eval_steps_per_second": 64.185,
+      "step": 3000
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 1.4861751152073732e-05,
+      "loss": 0.1315,
+      "step": 3050
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.1194,
+      "step": 3100
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 1.3709677419354839e-05,
+      "loss": 0.1237,
+      "step": 3150
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 1.313364055299539e-05,
+      "loss": 0.1183,
+      "step": 3200
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 1.2557603686635947e-05,
+      "loss": 0.1231,
+      "step": 3250
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 1.1981566820276497e-05,
+      "loss": 0.1209,
+      "step": 3300
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 1.1405529953917052e-05,
+      "loss": 0.1209,
+      "step": 3350
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 1.0829493087557604e-05,
+      "loss": 0.1213,
+      "step": 3400
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 1.0253456221198157e-05,
+      "loss": 0.1218,
+      "step": 3450
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.1163,
+      "step": 3500
+    },
+    {
+      "epoch": 8.06,
+      "eval_accuracy": 0.9556170518399595,
+      "eval_f1": 0.47168932038834954,
+      "eval_loss": 0.17665258049964905,
+      "eval_precision": 0.55633931843166,
+      "eval_recall": 0.4093973304570581,
+      "eval_runtime": 7.4971,
+      "eval_samples_per_second": 533.536,
+      "eval_steps_per_second": 66.692,
+      "step": 3500
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 9.101382488479262e-06,
+      "loss": 0.111,
+      "step": 3550
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 8.525345622119817e-06,
+      "loss": 0.112,
+      "step": 3600
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 7.949308755760369e-06,
+      "loss": 0.1106,
+      "step": 3650
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 7.373271889400922e-06,
+      "loss": 0.1098,
+      "step": 3700
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 6.7972350230414745e-06,
+      "loss": 0.1125,
+      "step": 3750
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 6.221198156682028e-06,
+      "loss": 0.1099,
+      "step": 3800
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 5.64516129032258e-06,
+      "loss": 0.1113,
+      "step": 3850
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5.0691244239631346e-06,
+      "loss": 0.1079,
+      "step": 3900
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 4.493087557603687e-06,
+      "loss": 0.1048,
+      "step": 3950
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 3.9170506912442395e-06,
+      "loss": 0.1045,
+      "step": 4000
+    },
+    {
+      "epoch": 9.22,
+      "eval_accuracy": 0.9554269297393215,
+      "eval_f1": 0.48804591584628837,
+      "eval_loss": 0.17828913033008575,
+      "eval_precision": 0.5594771241830065,
+      "eval_recall": 0.4327895375488742,
+      "eval_runtime": 7.85,
+      "eval_samples_per_second": 509.557,
+      "eval_steps_per_second": 63.695,
+      "step": 4000
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 3.341013824884793e-06,
+      "loss": 0.1054,
+      "step": 4050
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 2.7649769585253458e-06,
+      "loss": 0.1061,
+      "step": 4100
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 2.1889400921658987e-06,
+      "loss": 0.1008,
+      "step": 4150
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.1024,
+      "step": 4200
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 1.0368663594470047e-06,
+      "loss": 0.1034,
+      "step": 4250
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 4.6082949308755763e-07,
+      "loss": 0.1048,
+      "step": 4300
+    },
+    {
+      "epoch": 10.0,
+      "step": 4340,
+      "total_flos": 1.3513697075311152e+16,
+      "train_loss": 0.06244980893376786,
+      "train_runtime": 304.7861,
+      "train_samples_per_second": 1822.196,
+      "train_steps_per_second": 14.239
     }
   ],
   "logging_steps": 50,
-  "max_steps": 2170,
+  "max_steps": 4340,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 10,
   "save_steps": 500,
-  "total_flos": 6755106697897248.0,
+  "total_flos": 1.3513697075311152e+16,
   "train_batch_size": 128,
   "trial_name": null,
   "trial_params": null