Qwen 2 inference problem

[Sadeghi85]

I tried ExLlamaV2 of Qwen 2 7b, but the output is gibberish. There is a discussion over at llama.cpp: ggerganov/llama.cpp#7805

[Anthonyg5005]

Seems to be working fine for me. I'm using chat completions through tabbyAPI.
ExllamaV2 version: 0.1.4