Update src/RAGSample.py

src/RAGSample.py

@@ -370,19 +370,25 @@ Answer:
     
     # Initialize a local Hugging Face model
     hf_pipeline = pipeline(
+        # "text-generation",
+        # model="microsoft/BioGPT",
+        # tokenizer="microsoft/BioGPT",
+        # max_new_tokens=100,      # Reduced for stability
+        # max_length=1024,         # BioGPT's context length
+        # temperature=0.2,         # Lower for more focused responses
+        # device_map="auto",
+        # torch_dtype=torch.float16,
+        # return_full_text=False,
+        # truncation=True,
+        # do_sample=True,
+        # pad_token_id=1,
+        # eos_token_id=2,
         "text-generation",
         model="microsoft/BioGPT",
         tokenizer="microsoft/BioGPT",
-        max_new_tokens=100,      # Reduced for stability
-        max_length=1024,         # BioGPT's context length
-        temperature=0.2,         # Lower for more focused responses
+        max_new_tokens=50,  # Very small for testing
         device_map="auto",
-        torch_dtype=torch.float16,
-        return_full_text=False,
-        truncation=True,
-        do_sample=True,
-        pad_token_id=1,
-        eos_token_id=2,
+        torch_dtype=torch.float16
     )
     
     # Wrap it in LangChain