[GH-ISSUE #203] 因为需要迁移到310视频卡上，需要先转换成onnx格式,测试了多种方法都有技术阻碍 #165

New issue

Open

opened 2026-10-26 21:13:17 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 21:13:17 +00:00

Owner

Originally created by @furanger on GitHub (Mar 18, 2025).
Original GitHub issue: https://github.com/deepseek-ai/Janus/issues/203

一种是转换成 transformers、llama、swift格式，然后出现了
ValueError: The checkpoint you are trying to load has model type multi_modality but Transformers does not recognize this architecture ，看了一下应该是transformers 里没有配置文件中的multi_modality类别
第二种方法是onnxruntime,先启动模型，然后传输给onnxruntime,有了multi_modality类型了，
又报了
NotImplementedError: Module [MultiModalityCausalLM] is missing the required "forward" function
目前还没看懂MultiModalityCausalLM 类与AutoModelForCausalLM的forward应该如何设计，有什么好建议吗？

btw - 研化好像转换成了R1 但R1没有用MultiModalityCausalLM类

Originally created by @furanger on GitHub (Mar 18, 2025). Original GitHub issue: https://github.com/deepseek-ai/Janus/issues/203 一种是转换成 transformers、llama、swift格式，然后出现了 ValueError: The checkpoint you are trying to load has model type `multi_modality` but Transformers does not recognize this architecture ，看了一下应该是transformers 里没有配置文件中的multi_modality类别第二种方法是onnxruntime,先启动模型，然后传输给onnxruntime,有了multi_modality类型了，又报了 NotImplementedError: Module [MultiModalityCausalLM] is missing the required "forward" function 目前还没看懂MultiModalityCausalLM 类与AutoModelForCausalLM的forward应该如何设计，有什么好建议吗？ btw - 研化好像转换成了R1 但R1没有用MultiModalityCausalLM类

No labels

pull-request

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

deepseek-ai/Janus#165

No description provided.

Rows
Columns