Upload 28 files

models_bin/text_transformer/config.json

@@ -1,45 +1,51 @@
-{
-  "add_cross_attention": false,
-  "architectures": [
-    "BertForSequenceClassification"
-  ],
-  "attention_probs_dropout_prob": 0.1,
-  "bos_token_id": null,
-  "classifier_dropout": null,
-  "dtype": "float32",
-  "eos_token_id": null,
-  "gradient_checkpointing": false,
-  "hidden_act": "gelu",
-  "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
-  "id2label": {
-    "0": "LABEL_0",
-    "1": "LABEL_1",
-    "2": "LABEL_2",
-    "3": "LABEL_3",
-    "4": "LABEL_4"
-  },
-  "initializer_range": 0.02,
-  "intermediate_size": 3072,
-  "is_decoder": false,
-  "label2id": {
-    "LABEL_0": 0,
-    "LABEL_1": 1,
-    "LABEL_2": 2,
-    "LABEL_3": 3,
-    "LABEL_4": 4
-  },
-  "layer_norm_eps": 1e-12,
-  "max_position_embeddings": 512,
-  "model_type": "bert",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
-  "pad_token_id": 0,
-  "position_embedding_type": "absolute",
-  "problem_type": "single_label_classification",
-  "tie_word_embeddings": true,
-  "transformers_version": "5.13.1",
-  "type_vocab_size": 2,
-  "use_cache": false,
-  "vocab_size": 30522
-}
+{
+  "add_cross_attention": false,
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": null,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "eos_token_id": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "Normal",
+    "1": "Stress",
+    "2": "Depression",
+    "3": "Anxiety",
+    "4": "Emotional Distress"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "is_decoder": false,
+  "label2id": {
+    "Normal": 0,
+    "Stress": 1,
+    "Depression": 2,
+    "Anxiety": 3,
+    "Emotional Distress": 4
+  },
+    
+    
+    
+    
+    
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "tie_word_embeddings": true,
+  "transformers_version": "5.13.1",
+  "type_vocab_size": 2,
+  "use_cache": false,
+  "vocab_size": 30522
+}

src/models/text_classifier.py

@@ -177,37 +177,13 @@ class LinguisticStressClassifier:
         
         prob_dict = {res['label']: round(float(res['score']), 4) for res in results}
         
-        # FIX: The transformer was only trained on "Normal" and "Stress". 
-        # Its outputs for Depression, Anxiety, and Emotional Distress are random noise.
-        # We use heuristic logic to accurately detect these missing classes.
-        heuristic_res = self._heuristic_predict(text)
-        h_cat = heuristic_res["predicted_category"]
+        pred_category = max(prob_dict, key=prob_dict.get)
+        confidence = prob_dict[pred_category]
         
-        if h_cat in ["Depression", "Anxiety", "Emotional Distress"]:
-            # Override transformer noise with our accurate heuristic
-            prob_dict = heuristic_res["probabilities"]
-            pred_category = h_cat
-            confidence = heuristic_res["confidence"]
-            stress_score = heuristic_res["linguistic_stress_score"]
-        else:
-            # Zero out the noise for untrained classes
-            prob_dict["Depression"] = 0.0
-            prob_dict["Anxiety"] = 0.0
-            prob_dict["Emotional Distress"] = 0.0
-            
-            # Re-normalize Normal and Stress
-            total_valid = prob_dict.get("Normal", 0.0) + prob_dict.get("Stress", 0.0)
-            if total_valid > 0:
-                prob_dict["Normal"] = round(prob_dict["Normal"] / total_valid, 4)
-                prob_dict["Stress"] = round(prob_dict["Stress"] / total_valid, 4)
-            
-            pred_category = max(prob_dict, key=prob_dict.get)
-            confidence = prob_dict[pred_category]
-            
-            calm_prob = prob_dict.get("Normal", 0.0)
-            stress_prob = 1.0 - calm_prob
-            neg_density = calculate_negative_word_density(text)
-            stress_score = round(min(100.0, max(0.0, (stress_prob * 80.0) + (neg_density * 100.0))), 2)
+        calm_prob = prob_dict.get("Normal", 0.0)
+        stress_prob = 1.0 - calm_prob
+        neg_density = calculate_negative_word_density(text)
+        stress_score = round(min(100.0, max(0.0, (stress_prob * 80.0) + (neg_density * 100.0))), 2)
         
         return {
             "predicted_category": pred_category,