Upload folder using huggingface_hub

README.md

@@ -87,44 +87,65 @@ For **26 parent subjects**, F1-score improves to **0.934** with full metadata.
 from transformers import TextClassificationPipeline, pipeline
 import torch
 
+# Define the Custom Pipeline
 class ASJCMultiLabelPipeline(TextClassificationPipeline):
+    """
+    Custom pipeline for multi-label ASJC classification.
+    
+    This pipeline:
+      - Applies sigmoid to the model logits.
+      - Filters labels by a threshold.
+      - Returns all labels with scores above the threshold.
+    
+    Threshold can be specified during pipeline creation. 
+    If not provided, it defaults to the `threshold` in the model's config.json, or 0.3.
+    """
     def __init__(self, *args, **kwargs):
         self.threshold = kwargs.pop("threshold", None)
         super().__init__(*args, **kwargs)
 
-        # If no explicit threshold passed → use threshold from config.json
+        # Use threshold from config if none is passed explicitly
         if self.threshold is None:
             self.threshold = getattr(self.model.config, "threshold", 0.3)
 
     def postprocess(self, model_outputs, **kwargs):
+        # Convert logits to probabilities using sigmoid
         scores = torch.sigmoid(torch.tensor(model_outputs["logits"])).tolist()
         results = []
 
+        # Collect labels above the threshold
         for i, score in enumerate(scores):
             if score >= self.threshold:
                 label = self.model.config.id2label[str(i)]
                 results.append({"label": label, "score": float(score)})
 
-        # Sort by score descending
+        # Sort results by descending probability
         results = sorted(results, key=lambda x: x["score"], reverse=True)
         return results
 ```
 
 ```python
-pipe = pipeline("text-classification", model="asjc-classification/scibert_multilabel_asjc_classifier")
+# Create pipeline with the multi-label model
+pipe = pipeline(
+    "text-classification",
+    model="asjc-classification/scibert_multilabel_asjc_classifier"
+)
 
+# Example text input (title, container_title, abstract)
 text = (
     "title={Jodometrie}, "
     "container_title={Fresenius' Zeitschrift für analytische Chemie, Zeitschrift für analytische Chemie}, "
     "abstract={}"
 )
 
+# Get predictions
 result = pipe(text)
 print(result)
 
-# Actual labels: 
+# Expected labels (based on actual ASJC categories):
 # - Clinical Biochemistry
 # - Analytical Chemistry
+
 ```
 
 ---