Upload folder using huggingface_hub

fixtures/review.json

@@ -30,6 +30,36 @@
         ]
       }
     },
+    {
+      "start": 204,
+      "end": 216,
+      "text": "Battery life",
+      "label": "feature_mention",
+      "score": 0.3691289722919464,
+      "attributes": {
+        "sentiment": [
+          {
+            "label": "negative",
+            "score": 0.999858021736145
+          }
+        ]
+      }
+    },
+    {
+      "start": 260,
+      "end": 269,
+      "text": "fan noise",
+      "label": "feature_mention",
+      "score": 0.24698400497436523,
+      "attributes": {
+        "sentiment": [
+          {
+            "label": "negative",
+            "score": 0.9998291730880737
+          }
+        ]
+      }
+    },
     {
       "start": 341,
       "end": 349,
@@ -60,6 +90,21 @@
         ]
       }
     },
+    {
+      "start": 408,
+      "end": 419,
+      "text": "USB-C ports",
+      "label": "feature_mention",
+      "score": 0.4605136513710022,
+      "attributes": {
+        "sentiment": [
+          {
+            "label": "negative",
+            "score": 0.997340977191925
+          }
+        ]
+      }
+    },
     {
       "start": 451,
       "end": 458,

model.py

@@ -80,6 +80,7 @@ EXAMPLE_SCHEMAS = {
         },
     },
     "review": {
+        "threshold": 0.15,  # features in complaint clauses score below the 0.5 default
         "entities": {
             "feature_mention": "A feature, component, or aspect of a product mentioned in the review",
         },
@@ -210,7 +211,8 @@ def _infer_real(text):
     if attr_groups:
         schema = schema.entity_attributes(attr_groups)
 
-    result = _model.extract(text, schema, include_spans=True, include_confidence=True)
+    result = _model.extract(text, schema, threshold=cfg.get("threshold", 0.5),
+                            include_spans=True, include_confidence=True)
 
     spans = []
     for label, entries in result.get("entities", {}).items():