#!/bin/bash
lrank=$OMPI_COMM_WORLD_LOCAL_RANK
RANK=$OMPI_COMM_WORLD_RANK
WORLD_SIZE=$OMPI_COMM_WORLD_SIZE

DATA_PATH="./data"
CHECKPOINT_PATH="./checkpoint"

MICRO_BATCH_SIZE=1
GLOBAL_BATCH_SIZE=8


APP="python3 -u pretrain_vit.py \
       --num-layers 24 \
       --hidden-size 1024 \
       --num-attention-heads 16 \
       --micro-batch-size ${MICRO_BATCH_SIZE} \
       --global-batch-size ${GLOBAL_BATCH_SIZE} \
       --seq-length 1024 \
       --max-position-embeddings 1024 \
       --train-iters 500000 \
       --lr-decay-iters 320000 \
       --save $CHECKPOINT_PATH \
       --load $CHECKPOINT_PATH \
       --data-path $DATA_PATH \
       --data-impl mmap \
       --split 949,50,1 \
       --distributed-backend nccl \
       --lr 0.00015 \
       --min-lr 1.0e-5 \
       --lr-decay-style cosine \
       --weight-decay 1e-2 \
       --clip-grad 1.0 \
       --lr-warmup-fraction .01 \
       --checkpoint-activations \
       --log-interval 100 \
       --save-interval 10000 \
       --eval-interval 1000 \
       --eval-iters 10 \
       --fp16 \
       --padded_vocab_size 224\
       --rank $RANK \
       --world_size $WORLD_SIZE \
    "

# --eval-only True \
# --do_test True \

case ${lrank} in
[0])
  export HIP_VISIBLE_DEVICES=0,1,2,3
  export UCX_NET_DEVICES=mlx5_0:1
  export UCX_IB_PCI_BW=mlx5_0:50Gbs
  NCCL_SOCKET_IFNAME=ib0 numactl --cpunodebind=0 --membind=0 ${APP}
  ;;
[1])
  export HIP_VISIBLE_DEVICES=0,1,2,3
  export UCX_NET_DEVICES=mlx5_1:1
  export UCX_IB_PCI_BW=mlx5_1:50Gbs
  NCCL_SOCKET_IFNAME=ib0 numactl --cpunodebind=1 --membind=1 ${APP}
  ;;
[2])
  export HIP_VISIBLE_DEVICES=0,1,2,3
  export UCX_NET_DEVICES=mlx5_2:1
  export UCX_IB_PCI_BW=mlx5_2:50Gbs
  NCCL_SOCKET_IFNAME=ib0 numactl --cpunodebind=2 --membind=2 ${APP}
  ;;
[3])
  export HIP_VISIBLE_DEVICES=0,1,2,3
  export UCX_NET_DEVICES=mlx5_3:1
  export UCX_IB_PCI_BW=mlx5_3:50Gbs
  NCCL_SOCKET_IFNAME=ib0 numactl --cpunodebind=3 --membind=3 ${APP}
  ;;
esac