🔄 Incremental importance | Acc: 0.803, F1: 0.662

README.md

@@ -21,19 +21,19 @@ Fine-tuned FinBERT model for financial importance analysis in TradePulse.
 
 ## Performance
 
-*Last training: 2025-07-29 09:32*  
-*Dataset: `news_20250729.csv` (53 samples)*
+*Last training: 2025-07-29 13:33*  
+*Dataset: `base_reference.csv` (637 samples)*
 
 | Metric | Value |
 |--------|-------|
-| Loss | 2.5392 |
-| Accuracy | 0.6429 |
-| F1 Score | 0.5031 |
+| Loss | 0.6551 |
+| Accuracy | 0.8125 |
+| F1 Score | 0.8127 |
 
-| F1 Macro | 0.5031 |
+| F1 Macro | 0.8127 |
 
-| Precision | 0.4133 |
-| Recall | 0.6429 |
+| Precision | 0.8140 |
+| Recall | 0.8125 |
 
 ## Training Details
 
@@ -65,4 +65,4 @@ predictions = outputs.logits.softmax(dim=-1)
 ## Model Card Authors
 
 - TradePulse ML Team
-- Auto-generated on 2025-07-29 09:32:10
+- Auto-generated on 2025-07-29 13:33:51

checkpoint-20/trainer_state.json

@@ -1,86 +0,0 @@
-{
-  "best_metric": 0.5031055900621119,
-  "best_model_checkpoint": "hf-importance-production/checkpoint-20",
-  "epoch": 2.0,
-  "eval_steps": 500,
-  "global_step": 20,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 1.0,
-      "grad_norm": 181.4758758544922,
-      "learning_rate": 9.090909090909091e-06,
-      "loss": 3.3705,
-      "step": 10
-    },
-    {
-      "epoch": 1.0,
-      "eval_accuracy": 0.35714285714285715,
-      "eval_f1": 0.3383458646616541,
-      "eval_f1_macro": 0.17543859649122806,
-      "eval_loss": 1.8454712629318237,
-      "eval_precision": 0.32142857142857145,
-      "eval_precision_macro": 0.16666666666666666,
-      "eval_recall": 0.35714285714285715,
-      "eval_recall_macro": 0.1851851851851852,
-      "eval_runtime": 7.7174,
-      "eval_samples_per_second": 1.814,
-      "eval_steps_per_second": 0.518,
-      "step": 10
-    },
-    {
-      "epoch": 2.0,
-      "grad_norm": 83.5034408569336,
-      "learning_rate": 0.0,
-      "loss": 1.439,
-      "step": 20
-    },
-    {
-      "epoch": 2.0,
-      "eval_accuracy": 0.6428571428571429,
-      "eval_f1": 0.5031055900621119,
-      "eval_f1_macro": 0.2608695652173913,
-      "eval_loss": 2.539227247238159,
-      "eval_precision": 0.41326530612244905,
-      "eval_precision_macro": 0.2142857142857143,
-      "eval_recall": 0.6428571428571429,
-      "eval_recall_macro": 0.3333333333333333,
-      "eval_runtime": 7.725,
-      "eval_samples_per_second": 1.812,
-      "eval_steps_per_second": 0.518,
-      "step": 20
-    }
-  ],
-  "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": true
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 20522846582784.0,
-  "train_batch_size": 4,
-  "trial_name": null,
-  "trial_params": null
-}

{checkpoint-20 → checkpoint-240}/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9451c2dc524d02e41d47d82c6085a255f1022b0b351a2d0a6fb9dd8386e8ff77
+oid sha256:5adf2aaa4fe85784e8bc042af10862ff3a302c4a368f116c0a5f33220511fc72
 size 439039996

checkpoint-240/trainer_state.json

@@ -0,0 +1,240 @@
+{
+  "best_metric": 0.8126807447116778,
+  "best_model_checkpoint": "hf-importance-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": 70.0009765625,
+      "learning_rate": 8.403361344537816e-07,
+      "loss": 0.7483,
+      "step": 10
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 24.547622680664062,
+      "learning_rate": 1.6806722689075632e-06,
+      "loss": 0.9913,
+      "step": 20
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 57.52896499633789,
+      "learning_rate": 2.521008403361345e-06,
+      "loss": 0.6966,
+      "step": 30
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 74.12036895751953,
+      "learning_rate": 3.3613445378151265e-06,
+      "loss": 0.6548,
+      "step": 40
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 9.066349983215332,
+      "learning_rate": 4.201680672268908e-06,
+      "loss": 0.3729,
+      "step": 50
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 50.364627838134766,
+      "learning_rate": 5.04201680672269e-06,
+      "loss": 0.6898,
+      "step": 60
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 78.47008514404297,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.7254,
+      "step": 70
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.9752311706542969,
+      "learning_rate": 6.722689075630253e-06,
+      "loss": 0.6388,
+      "step": 80
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.438714027404785,
+      "learning_rate": 7.563025210084034e-06,
+      "loss": 0.5719,
+      "step": 90
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 70.23839569091797,
+      "learning_rate": 8.403361344537815e-06,
+      "loss": 0.7578,
+      "step": 100
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 85.30646514892578,
+      "learning_rate": 9.243697478991598e-06,
+      "loss": 1.2087,
+      "step": 110
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3721972703933716,
+      "learning_rate": 9.917355371900828e-06,
+      "loss": 0.7927,
+      "step": 120
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8125,
+      "eval_f1": 0.8083101323053494,
+      "eval_f1_macro": 0.6997588679484738,
+      "eval_loss": 0.6532053351402283,
+      "eval_precision": 0.8070379283788987,
+      "eval_precision_macro": 0.7347195481966371,
+      "eval_recall": 0.8125,
+      "eval_recall_macro": 0.6755615453728661,
+      "eval_runtime": 88.1395,
+      "eval_samples_per_second": 1.815,
+      "eval_steps_per_second": 0.454,
+      "step": 120
+    },
+    {
+      "epoch": 1.0833333333333333,
+      "grad_norm": 53.19272232055664,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.347,
+      "step": 130
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 130.35791015625,
+      "learning_rate": 8.264462809917356e-06,
+      "loss": 0.6841,
+      "step": 140
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 24.545459747314453,
+      "learning_rate": 7.43801652892562e-06,
+      "loss": 0.5361,
+      "step": 150
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 11.489201545715332,
+      "learning_rate": 6.611570247933885e-06,
+      "loss": 0.1774,
+      "step": 160
+    },
+    {
+      "epoch": 1.4166666666666667,
+      "grad_norm": 7.724330902099609,
+      "learning_rate": 5.785123966942149e-06,
+      "loss": 0.1986,
+      "step": 170
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 51.44404602050781,
+      "learning_rate": 4.958677685950414e-06,
+      "loss": 0.2807,
+      "step": 180
+    },
+    {
+      "epoch": 1.5833333333333335,
+      "grad_norm": 58.367225646972656,
+      "learning_rate": 4.132231404958678e-06,
+      "loss": 0.5279,
+      "step": 190
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.692018985748291,
+      "learning_rate": 3.3057851239669424e-06,
+      "loss": 0.3302,
+      "step": 200
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 39.534156799316406,
+      "learning_rate": 2.479338842975207e-06,
+      "loss": 1.0246,
+      "step": 210
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 45.754249572753906,
+      "learning_rate": 1.6528925619834712e-06,
+      "loss": 0.6398,
+      "step": 220
+    },
+    {
+      "epoch": 1.9166666666666665,
+      "grad_norm": 55.728824615478516,
+      "learning_rate": 8.264462809917356e-07,
+      "loss": 0.3741,
+      "step": 230
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.00028416712302714586,
+      "learning_rate": 0.0,
+      "loss": 0.4744,
+      "step": 240
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8125,
+      "eval_f1": 0.8126807447116778,
+      "eval_f1_macro": 0.6772336520815223,
+      "eval_loss": 0.6551325917243958,
+      "eval_precision": 0.8140347917079452,
+      "eval_precision_macro": 0.6750957664619245,
+      "eval_recall": 0.8125,
+      "eval_recall_macro": 0.6814734950584006,
+      "eval_runtime": 88.2081,
+      "eval_samples_per_second": 1.814,
+      "eval_steps_per_second": 0.453,
+      "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 sha256:9451c2dc524d02e41d47d82c6085a255f1022b0b351a2d0a6fb9dd8386e8ff77
+oid sha256:5adf2aaa4fe85784e8bc042af10862ff3a302c4a368f116c0a5f33220511fc72
 size 439039996

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4d854fcf3b8b50fb52ab209646977147211be9331153446e6dbf38c9cb0bb27f
+oid sha256:e92fee1fb04aadd9214a2ac739f02f3f2f13b9a2c138bafa3f4c239de58adc1b
 size 5521