[GH-ISSUE #67] 请问model_length设置为2048进行微调，微调后模型输入最大只能是2048了吗？ #57

New issue

Closed

opened 2026-10-26 21:02:53 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 21:02:53 +00:00

Owner

Originally created by @Stucrehap on GitHub (Dec 2, 2023).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-Coder/issues/67

model_length设置为2048进行微调，使用微调后的模型推理，输入长度超过2048后，因为输入超长而报错。
将tokenizer的model_length设置为8196后，推理时超长输入未报错，但是模型不输出新的token。
请问是什么原因呢

Originally created by @Stucrehap on GitHub (Dec 2, 2023). Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-Coder/issues/67 model_length设置为2048进行微调，使用微调后的模型推理，输入长度超过2048后，因为输入超长而报错。将tokenizer的model_length设置为8196后，推理时超长输入未报错，但是模型不输出新的token。请问是什么原因呢

bsakaguchi closed this issue

2026-10-26 21:02:53 +00:00

bsakaguchi referenced this issue

2026-10-26 21:05:01 +00:00

[PR #57] [MERGED] Update README.md #249

bsakaguchi referenced this issue

2026-10-26 21:05:13 +00:00

[PR #634] [CLOSED] Fix the position of add_generation_prompt #272

bsakaguchi referenced this issue

2026-10-26 21:05:28 +00:00

[PR #692] Add GitHub Actions workflow for Python package using Conda #290