The expanded size of the tensor (472) must match the existing size (1023) at non-singleton dimension 1.

[lafreak]

Self Checks

• [X] This is only for bug report, if you would like to ask a question, please head to Discussions.
• [X] I have searched for existing issues search for existing issues, including closed ones.
• [X] I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [X] [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• [X] Please do not modify this template :) and fill in all the required fields.

Cloud or Self Hosted

Self Hosted (Source)

Steps to reproduce

Infer via WebUI with mid-size prompt (5-6 sentences)

Configs default: Iterative Prompt Length = 200 Maximum tokens per batch = 1024 Top-P 0.7 Repetition Penalty 1.2 Temperature 0.7

Model: default one, from: https://speech.fish.audio/

✔️ Expected Behavior

Audio being generated

❌ Actual Behavior

The expanded size of the tensor (895) must match the existing size (1023) at non-singleton dimension 1. Target sizes: [9, 895]. Tensor sizes: [9, 1023]