Update processing.py

processing.py

@@ -59,7 +59,7 @@ class SequentialAnalyzer:
         tokenizer = AutoTokenizer.from_pretrained(model_name, use_auth_token=self.hf_token)
         model = AutoModelForCausalLM.from_pretrained(
             model_name,
-            torch_dtype=torch.bfloat16,
+            torch_dtype=torch.float16,
             device_map="auto",
             use_auth_token=self.hf_token
         )
@@ -75,10 +75,10 @@ class SequentialAnalyzer:
             "text-generation",
             model=model,
             tokenizer=tokenizer,
-            max_new_tokens=512,
+            max_new_tokens=400,
             temperature=0.4,
             top_p=0.9,
-            top_k=50,
+            top_k=35,
             repetition_penalty=1.2,
         )
 
@@ -102,7 +102,6 @@ class SequentialAnalyzer:
             input_variables=["context", "question"]
         )
 
-    @spaces.GPU(duration=600)  # Increased duration to account for all tasks
     def analyze_content(self, content, safe_progress):
         safe_progress(0.6, desc="Analyzing attachments...")
         attachments_answer = self.analyze_task(content, attachments_task, attachments_db)