End of training

README.md

@@ -3,6 +3,8 @@ license: apache-2.0
 base_model: bert-base-uncased
 tags:
 - generated_from_trainer
+datasets:
+- fursov/gec_ner_val3
 metrics:
 - precision
 - recall
@@ -10,7 +12,26 @@ metrics:
 - accuracy
 model-index:
 - name: ner-gec-v2
-  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: 0.36697832554186144
+    - name: Recall
+      type: recall
+      value: 0.23284346770931644
+    - name: F1
+      type: f1
+      value: 0.2849129753361379
+    - name: Accuracy
+      type: accuracy
+      value: 0.941991634627572
 ---
 
 <!-- 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-v2
 
-This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
+This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the fursov/gec_ner_val3 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2137
-- Precision: 0.3467
-- Recall: 0.2212
-- F1: 0.2701
-- Accuracy: 0.9399
+- Loss: 0.2067
+- Precision: 0.3670
+- Recall: 0.2328
+- F1: 0.2849
+- Accuracy: 0.9420
 
 ## Model description
 

all_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.941991634627572,
+    "eval_f1": 0.2849129753361379,
+    "eval_loss": 0.20673920214176178,
+    "eval_precision": 0.36697832554186144,
+    "eval_recall": 0.23284346770931644,
+    "eval_runtime": 7.4594,
+    "eval_samples": 4000,
+    "eval_samples_per_second": 536.235,
+    "eval_steps_per_second": 67.029,
+    "train_loss": 0.22592788950730397,
+    "train_runtime": 193.8299,
+    "train_samples": 55538,
+    "train_samples_per_second": 859.589,
+    "train_steps_per_second": 6.717
+}

eval_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.941991634627572,
+    "eval_f1": 0.2849129753361379,
+    "eval_loss": 0.20673920214176178,
+    "eval_precision": 0.36697832554186144,
+    "eval_recall": 0.23284346770931644,
+    "eval_runtime": 7.4594,
+    "eval_samples": 4000,
+    "eval_samples_per_second": 536.235,
+    "eval_steps_per_second": 67.029
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "train_loss": 0.22592788950730397,
+    "train_runtime": 193.8299,
+    "train_samples": 55538,
+    "train_samples_per_second": 859.589,
+    "train_steps_per_second": 6.717
+}

trainer_state.json

@@ -0,0 +1,210 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1302,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.807987711213518e-05,
+      "loss": 0.3238,
+      "step": 50
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6159754224270356e-05,
+      "loss": 0.2791,
+      "step": 100
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.423963133640553e-05,
+      "loss": 0.2721,
+      "step": 150
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.231950844854071e-05,
+      "loss": 0.2606,
+      "step": 200
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0399385560675886e-05,
+      "loss": 0.2566,
+      "step": 250
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.847926267281106e-05,
+      "loss": 0.2511,
+      "step": 300
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.655913978494624e-05,
+      "loss": 0.2499,
+      "step": 350
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4639016897081416e-05,
+      "loss": 0.25,
+      "step": 400
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.271889400921659e-05,
+      "loss": 0.2391,
+      "step": 450
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.079877112135177e-05,
+      "loss": 0.2324,
+      "step": 500
+    },
+    {
+      "epoch": 1.15,
+      "eval_accuracy": 0.9352739870716972,
+      "eval_f1": 0.12380682355164917,
+      "eval_loss": 0.23587752878665924,
+      "eval_precision": 0.2070164348925411,
+      "eval_recall": 0.08831063772414723,
+      "eval_runtime": 7.6555,
+      "eval_samples_per_second": 522.501,
+      "eval_steps_per_second": 65.313,
+      "step": 500
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.8878648233486942e-05,
+      "loss": 0.2283,
+      "step": 550
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.6958525345622122e-05,
+      "loss": 0.2233,
+      "step": 600
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.50384024577573e-05,
+      "loss": 0.2256,
+      "step": 650
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3118279569892472e-05,
+      "loss": 0.221,
+      "step": 700
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1198156682027652e-05,
+      "loss": 0.22,
+      "step": 750
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.927803379416283e-05,
+      "loss": 0.2202,
+      "step": 800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7357910906298002e-05,
+      "loss": 0.2124,
+      "step": 850
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.543778801843318e-05,
+      "loss": 0.2009,
+      "step": 900
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3517665130568357e-05,
+      "loss": 0.1915,
+      "step": 950
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1597542242703534e-05,
+      "loss": 0.1901,
+      "step": 1000
+    },
+    {
+      "epoch": 2.3,
+      "eval_accuracy": 0.9398686045037813,
+      "eval_f1": 0.2701234567901235,
+      "eval_loss": 0.21368014812469482,
+      "eval_precision": 0.34671455736319456,
+      "eval_recall": 0.22124848321423757,
+      "eval_runtime": 7.6592,
+      "eval_samples_per_second": 522.246,
+      "eval_steps_per_second": 65.281,
+      "step": 1000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.1894,
+      "step": 1050
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.757296466973887e-06,
+      "loss": 0.188,
+      "step": 1100
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.837173579109064e-06,
+      "loss": 0.1886,
+      "step": 1150
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.9170506912442395e-06,
+      "loss": 0.1874,
+      "step": 1200
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.996927803379416e-06,
+      "loss": 0.1888,
+      "step": 1250
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 7.680491551459294e-08,
+      "loss": 0.1859,
+      "step": 1300
+    },
+    {
+      "epoch": 3.0,
+      "step": 1302,
+      "total_flos": 4487551835152560.0,
+      "train_loss": 0.22592788950730397,
+      "train_runtime": 193.8299,
+      "train_samples_per_second": 859.589,
+      "train_steps_per_second": 6.717
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 1302,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 4487551835152560.0,
+  "train_batch_size": 128,
+  "trial_name": null,
+  "trial_params": null
+}