moonshotai

/

Moonlight-16B-A3B-Instruct

like 80

Follow

Moonshot AI 120

Fix cache API call: Replace get_max_length() with get_seq_length()

Description:
Updated the prepare_inputs_for_generation method in modeling_deepseek.py to use get_seq_length() instead of the deprecated get_max_length(). This fixes an AttributeError caused by recent changes in the Transformers library.