You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Please pip install mamba-ssm causal-conv1d>=1.2.0
and set --micro_train_batch_size 1 to reproduce this bug ( IMPORTANT! ) --micro_train_batch_size 4 works well
Model Link: https://huggingface.co/ai21labs/Jamba-v0.1
Reproduce script in OpenRLHF: https://github.com/OpenLLMAI/OpenRLHF/blob/main/examples/scripts/train_sft_jamba_lora.sh
Please
pip install mamba-ssm causal-conv1d>=1.2.0
and set
--micro_train_batch_size 1
to reproduce this bug ( IMPORTANT! )--micro_train_batch_size 4
works wellMixtral + ZeRO3 + LoRA works well with the same hyperparameters (`--micro_train_batch_size 1):
see https://github.com/OpenLLMAI/OpenRLHF/blob/main/examples/scripts/train_sft_mixtral_lora.sh
The text was updated successfully, but these errors were encountered: