13 lines
424 B
Bash
13 lines
424 B
Bash
source bigdl-llm-init -t -g
|
|
export MASTER_ADDR=127.0.0.1
|
|
export CCL_ZE_IPC_EXCHANGE=sockets
|
|
NUM_GPUS=4
|
|
if [[ -n $OMP_NUM_THREADS ]]; then
|
|
export OMP_NUM_THREADS=$(($OMP_NUM_THREADS / $NUM_GPUS))
|
|
else
|
|
export OMP_NUM_THREADS=$(($(nproc) / $NUM_GPUS))
|
|
fi
|
|
torchrun --standalone \
|
|
--nnodes=1 \
|
|
--nproc-per-node $NUM_GPUS \
|
|
deepspeed_autotp.py --repo-id-or-model-path "meta-llama/Llama-2-7b-hf"
|