mirror of
https://github.com/hiyouga/LLaMA-Factory.git
synced 2025-08-02 03:32:50 +08:00
7 lines
241 B
Bash
7 lines
241 B
Bash
#!/bin/bash
|
|
# also launch it on slave machine using slave_config.yaml
|
|
|
|
CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch \
|
|
--config_file examples/accelerate/master_config.yaml \
|
|
src/train.py examples/lora_multi_gpu/llama3_lora_sft.yaml
|