set max tokens to 5k

This commit is contained in:
Dobromir Popov
2026-07-07 18:21:13 +02:00
parent aba5fb12fa
commit 456c43ea1d
3 changed files with 4 additions and 4 deletions

View File

@@ -321,7 +321,7 @@ class _TorchHandler(http.server.BaseHTTPRequestHandler):
if backend is None or not backend.is_head:
self._send_json(400, {"error": "model not loaded on this node"})
return
max_tokens = int(body.get("max_tokens") or body.get("max_new_tokens") or 256)
max_tokens = int(body.get("max_tokens") or body.get("max_new_tokens") or 5120)
temperature = float(body.get("temperature") or 1.0)
top_p = float(body.get("top_p") or 1.0)