Upload 22 files

app.py

@@ -306,9 +306,9 @@ def _run_text_inference_with_model(primary_model_id: str, prompt: str, max_token
                 "temperature": NVIDIA_TEMPERATURE,
                 "top_p": NVIDIA_TOP_P,
                 "stream": True,
-                "include_reasoning": False,
-                "chat_template_kwargs": {"enable_thinking": NVIDIA_ENABLE_THINKING},
             }
+            if NVIDIA_ENABLE_THINKING:
+                payload["chat_template_kwargs"] = {"enable_thinking": True}
             headers = {
                 "Authorization": f"Bearer {api_key}",
                 "Accept": "text/event-stream",
@@ -2051,9 +2051,9 @@ def run_nvidia_inference(image: Image.Image, prompt: str, max_tokens: int = NVID
                 "temperature": NVIDIA_TEMPERATURE,
                 "top_p": NVIDIA_TOP_P,
                 "stream": True,
-                "include_reasoning": False,
-                "chat_template_kwargs": {"enable_thinking": NVIDIA_ENABLE_THINKING},
             }
+            if NVIDIA_ENABLE_THINKING:
+                payload["chat_template_kwargs"] = {"enable_thinking": True}
             headers = {
                 "Authorization": f"Bearer {api_key}",
                 "Accept": "text/event-stream",
@@ -2139,9 +2139,9 @@ def run_nvidia_text_inference(prompt: str, max_tokens: int = OUTFIT_TEXT_SELECTO
                 "temperature": NVIDIA_TEMPERATURE,
                 "top_p": NVIDIA_TOP_P,
                 "stream": True,
-                "include_reasoning": False,
-                "chat_template_kwargs": {"enable_thinking": NVIDIA_ENABLE_THINKING},
             }
+            if NVIDIA_ENABLE_THINKING:
+                payload["chat_template_kwargs"] = {"enable_thinking": True}
             headers = {
                 "Authorization": f"Bearer {api_key}",
                 "Accept": "text/event-stream",