gpt-omni / mini-omni

open-source multimodal large language model that can hear, talk while thinking. Featuring real-time end-to-end speech input and streaming audio output conversational capabilities.
https://arxiv.org/abs/2408.16725
MIT License
3.06k stars 273 forks source link

运行server.py,进度条只加载到4%,这正常吗 #62

Closed YMMF007 closed 1 month ago

YMMF007 commented 1 month ago

image

Evgeny105 commented 1 month ago

I added to inference.py: torch.set_float32_matmul_precision("high")

YMMF007 commented 1 month ago

Thanks!

superFilicos commented 1 month ago

正常的,进度条的总长度是推理的最长,默认2048,不过扩展到最大也无所谓,进度条结束的长度就是实际有多少组token就说完啦

YMMF007 commented 1 month ago

噢噢噢噢,谢谢~

------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2024年9月10日(星期二) 下午2:26 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [gpt-omni/mini-omni] 运行server.py,进度条只加载到4%,这正常吗 (Issue #62)

正常的,进度条的总长度是推理的最长,默认2048,不过扩展到最大也无所谓,进度条结束的长度就是实际有多少组token就说完啦

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>