🔄 Incremental label | Acc: 0.845, F1: 0.836

README.md

@@ -21,19 +21,19 @@ Fine-tuned FinBERT model for financial sentiment analysis in TradePulse.
 
 ## Performance
 
-*Last training: 2025-07-29 09:27*  
-*Dataset: `news_20250729.csv` (53 samples)*
+*Last training: 2025-07-29 12:46*  
+*Dataset: `base_reference.csv` (637 samples)*
 
 | Metric | Value |
 |--------|-------|
-| Loss | 2.1347 |
-| Accuracy | 0.5714 |
-| F1 Score | 0.5536 |
+| Loss | 0.9526 |
+| Accuracy | 0.8250 |
+| F1 Score | 0.8217 |
 
-| F1 Macro | 0.5536 |
+| F1 Macro | 0.8217 |
 
-| Precision | 0.6429 |
-| Recall | 0.5714 |
+| Precision | 0.8277 |
+| Recall | 0.8250 |
 
 ## Training Details
 
@@ -65,4 +65,4 @@ predictions = outputs.logits.softmax(dim=-1)
 ## Model Card Authors
 
 - TradePulse ML Team
-- Auto-generated on 2025-07-29 09:27:33
+- Auto-generated on 2025-07-29 12:46:43

checkpoint-10/trainer_state.json

@@ -1,64 +0,0 @@
-{
-  "best_metric": 0.5535714285714286,
-  "best_model_checkpoint": "hf-sentiment-production/checkpoint-10",
-  "epoch": 1.0,
-  "eval_steps": 500,
-  "global_step": 10,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 1.0,
-      "grad_norm": 116.80064392089844,
-      "learning_rate": 9.090909090909091e-06,
-      "loss": 2.1666,
-      "step": 10
-    },
-    {
-      "epoch": 1.0,
-      "eval_accuracy": 0.5714285714285714,
-      "eval_f1": 0.5535714285714286,
-      "eval_f1_macro": 0.5787037037037037,
-      "eval_loss": 2.1347310543060303,
-      "eval_precision": 0.6428571428571429,
-      "eval_precision_macro": 0.7222222222222222,
-      "eval_recall": 0.5714285714285714,
-      "eval_recall_macro": 0.5555555555555555,
-      "eval_runtime": 7.7629,
-      "eval_samples_per_second": 1.803,
-      "eval_steps_per_second": 0.515,
-      "step": 10
-    }
-  ],
-  "logging_steps": 10,
-  "max_steps": 20,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
-  "save_steps": 500,
-  "stateful_callbacks": {
-    "EarlyStoppingCallback": {
-      "args": {
-        "early_stopping_patience": 1,
-        "early_stopping_threshold": 0.0
-      },
-      "attributes": {
-        "early_stopping_patience_counter": 0
-      }
-    },
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": false
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 10261423291392.0,
-  "train_batch_size": 4,
-  "trial_name": null,
-  "trial_params": null
-}

{checkpoint-10 → checkpoint-240}/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:762e4b5f3de0d6932bc387b2ca47fab34971ea86f950d431db351f7d5269f6ba
+oid sha256:f2b8d9dbc3c718fdf0698f3013f7f79c7162a22a441338d87fa34509632536c6
 size 439039996

checkpoint-240/trainer_state.json

@@ -0,0 +1,240 @@
+{
+  "best_metric": 0.8217223565907776,
+  "best_model_checkpoint": "hf-sentiment-production/checkpoint-240",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 240,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 32.42210006713867,
+      "learning_rate": 8.403361344537816e-07,
+      "loss": 0.5341,
+      "step": 10
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 32.30667495727539,
+      "learning_rate": 1.6806722689075632e-06,
+      "loss": 0.2982,
+      "step": 20
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 65.03559112548828,
+      "learning_rate": 2.521008403361345e-06,
+      "loss": 0.4933,
+      "step": 30
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 11.74012565612793,
+      "learning_rate": 3.3613445378151265e-06,
+      "loss": 0.3533,
+      "step": 40
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 88.07356262207031,
+      "learning_rate": 4.201680672268908e-06,
+      "loss": 0.9541,
+      "step": 50
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 56.444908142089844,
+      "learning_rate": 5.04201680672269e-06,
+      "loss": 0.857,
+      "step": 60
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 7.758561611175537,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.8653,
+      "step": 70
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 2.285611629486084,
+      "learning_rate": 6.722689075630253e-06,
+      "loss": 1.3253,
+      "step": 80
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 51.23533630371094,
+      "learning_rate": 7.563025210084034e-06,
+      "loss": 0.6127,
+      "step": 90
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 79.65983581542969,
+      "learning_rate": 8.403361344537815e-06,
+      "loss": 0.6925,
+      "step": 100
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 98.18861389160156,
+      "learning_rate": 9.243697478991598e-06,
+      "loss": 0.7733,
+      "step": 110
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.007537131663411856,
+      "learning_rate": 9.917355371900828e-06,
+      "loss": 0.3243,
+      "step": 120
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.81875,
+      "eval_f1": 0.8149920382165605,
+      "eval_f1_macro": 0.8011677282377919,
+      "eval_loss": 0.8371865153312683,
+      "eval_precision": 0.8279807278208441,
+      "eval_precision_macro": 0.8436405397645709,
+      "eval_recall": 0.81875,
+      "eval_recall_macro": 0.7802058251146221,
+      "eval_runtime": 88.4094,
+      "eval_samples_per_second": 1.81,
+      "eval_steps_per_second": 0.452,
+      "step": 120
+    },
+    {
+      "epoch": 1.0833333333333333,
+      "grad_norm": 77.50802612304688,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.4296,
+      "step": 130
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 15.326767921447754,
+      "learning_rate": 8.264462809917356e-06,
+      "loss": 0.1833,
+      "step": 140
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 91.35188293457031,
+      "learning_rate": 7.43801652892562e-06,
+      "loss": 0.3804,
+      "step": 150
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 88.58149719238281,
+      "learning_rate": 6.611570247933885e-06,
+      "loss": 0.5602,
+      "step": 160
+    },
+    {
+      "epoch": 1.4166666666666667,
+      "grad_norm": 44.488040924072266,
+      "learning_rate": 5.785123966942149e-06,
+      "loss": 0.1753,
+      "step": 170
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 71.1673583984375,
+      "learning_rate": 4.958677685950414e-06,
+      "loss": 0.2846,
+      "step": 180
+    },
+    {
+      "epoch": 1.5833333333333335,
+      "grad_norm": 68.90252685546875,
+      "learning_rate": 4.132231404958678e-06,
+      "loss": 0.3893,
+      "step": 190
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 9.828429222106934,
+      "learning_rate": 3.3057851239669424e-06,
+      "loss": 0.2814,
+      "step": 200
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 29.61725616455078,
+      "learning_rate": 2.479338842975207e-06,
+      "loss": 0.3129,
+      "step": 210
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 108.5630111694336,
+      "learning_rate": 1.6528925619834712e-06,
+      "loss": 0.3268,
+      "step": 220
+    },
+    {
+      "epoch": 1.9166666666666665,
+      "grad_norm": 21.261966705322266,
+      "learning_rate": 8.264462809917356e-07,
+      "loss": 0.8541,
+      "step": 230
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0011018794029951096,
+      "learning_rate": 0.0,
+      "loss": 0.8383,
+      "step": 240
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.825,
+      "eval_f1": 0.8217223565907776,
+      "eval_f1_macro": 0.8096114411903885,
+      "eval_loss": 0.9526222348213196,
+      "eval_precision": 0.8277439948473138,
+      "eval_precision_macro": 0.8298729509231896,
+      "eval_recall": 0.825,
+      "eval_recall_macro": 0.798550868049829,
+      "eval_runtime": 88.1644,
+      "eval_samples_per_second": 1.815,
+      "eval_steps_per_second": 0.454,
+      "step": 240
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 240,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 1,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 251010200512512.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:762e4b5f3de0d6932bc387b2ca47fab34971ea86f950d431db351f7d5269f6ba
+oid sha256:f2b8d9dbc3c718fdf0698f3013f7f79c7162a22a441338d87fa34509632536c6
 size 439039996

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6b8f159013fb904e2f36d2e95adb0ed11cf2cbd3e2b66c3ee3ec1d0c1082226e
+oid sha256:76b8cf3932f11a5314ed3d8475b3e07f076b574cf6ed23ea9ee7b8a1833b39a3
 size 5521