#1815 the reply repeated one token ('!') 256 times in a row:

Description

NVIDIA-SMI 595.91.07              Driver Version: 595.91.07      CUDA Version: 13.2   
NVIDIA GeForce RTX 3090       
35%   47C    P8             26W /  300W |   23874MiB /  24576MiB 
./setup.sh --host 0.0.0.0  --port 11500 --models-dir  ../Strata-data/unsloth_Qwen3.8-Flash-Next-GGUF --family  unsloth --model UD-Q4_K_XL
after start success

try say hi on web,only got !!!!!!!
the reply repeated one token ('!') 256 times in a row: ended as "length" (repeat_stop_tokens in strata-<model>.json; 0 turns this off). If a new request with a short prompt does the same, restart the server and report it (#606)
[strata] done: 256 tokens in 6 s (58.0 tok/s) (length, cancel=False), expert cache 96.7% hit (+0.5% of the routed experts over PCIe)
[strata] the reply reached max tokens while still thinking, so it has no answer: a thinking budget (reasoning_budget_tokens, in the request or in strata-<model>.json for every request) leaves room to answer