MedicalGPT: Training Your Own Medical GPT Model with ChatGPT Training Pipeline. 训练医疗大模型,实现了包括增量预训练(PT)、有监督微调(SFT)、RLHF、DPO、ORPO。
2.94k
stars
451
forks
source link
运行inference.py文件,报AttributeError: property 'eos_token' of 'ChatGLMTokenizer' object has no setter #353
Closed
liulint closed 3 months ago