Spaces:
Paused
Paused
P7.3 LLM-Fallback mit Nutzer-Token
Browse files- server/main.py +1 -1
server/main.py
CHANGED
|
@@ -464,7 +464,7 @@ async def llm_ep(req: Request, prompt: str = Form(...), system: str = Form(""),
|
|
| 464 |
try:
|
| 465 |
from gradio_client import Client
|
| 466 |
def g():
|
| 467 |
-
c = Client("huggingface-projects/llama-3.2-3B-Instruct", verbose=False)
|
| 468 |
return c.predict((system + "\n\n" if system else "") + prompt, min(mt, 2048), 0.3, 0.9, 50, 1.1, api_name="/generate")
|
| 469 |
out = await asyncio.to_thread(g)
|
| 470 |
return {"text": out if isinstance(out, str) else json.dumps(out), "model": "huggingface-projects/llama-3.2-3B-Instruct", "note": err}
|
|
|
|
| 464 |
try:
|
| 465 |
from gradio_client import Client
|
| 466 |
def g():
|
| 467 |
+
c = Client("huggingface-projects/llama-3.2-3B-Instruct", hf_token=(s or {}).get("token"), verbose=False)
|
| 468 |
return c.predict((system + "\n\n" if system else "") + prompt, min(mt, 2048), 0.3, 0.9, 50, 1.1, api_name="/generate")
|
| 469 |
out = await asyncio.to_thread(g)
|
| 470 |
return {"text": out if isinstance(out, str) else json.dumps(out), "model": "huggingface-projects/llama-3.2-3B-Instruct", "note": err}
|