UserWarning: None of the inputs have requires_grad=True. Gradients will be None

shibing624 / MedicalGPT

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。

Apache License 2.0

2.94k stars 451 forks source link

Closed cove1011 closed 2 months ago

cove1011 commented 2 months ago

学习率为零，后面的chosen和rejected全为none

cove1011 commented 2 months ago

Could not estimate the number of tokens of the input, floating-point operations will not be computed

shibing624 commented 2 months ago

use float32.