refactor evaluation

This commit is contained in:
YeAnbang
2024-07-22 05:57:39 +00:00
parent c5f582f666
commit 12fe8b5858
22 changed files with 309 additions and 1388 deletions

View File

@@ -17,6 +17,7 @@ set_n_least_used_CUDA_VISIBLE_DEVICES 4
PROJECT_NAME="SFT"
PARENT_SAVE_DIR="" # Path to a folder to save checkpoints
PARENT_CONFIG_FILE="" # Path to a folder to save training config logs
PARENT_LOG_DIR="" # Path to a folder to save training config logs
PRETRAINED_MODEL_PATH="" # huggingface or local model path
PRETRAINED_TOKENIZER_PATH="" # huggingface or local tokenizer path
declare -a dataset=(
@@ -36,6 +37,7 @@ TIMESTAMP=$(date +%Y-%m-%d-%H-%M-%S)
FULL_PROJECT_NAME="${PROJECT_NAME}-${TIMESTAMP}"
SAVE_DIR="${PARENT_SAVE_DIR}${FULL_PROJECT_NAME}"
CONFIG_FILE="${PARENT_CONFIG_FILE}${FULL_PROJECT_NAME}.json"
LOG_DIR="${PARENT_LOG_DIR}${FULL_PROJECT_NAME}"
echo $(which colossalai)
echo $(which python)
@@ -47,6 +49,7 @@ colossalai run --nproc_per_node 4 --master_port 31312 --hostfile ./hostfile trai
--dataset ${dataset[@]} \
--save_path $SAVE_DIR \
--config_file $CONFIG_FILE \
--log_dir $LOG_DIR \
--lora_rank 0 \
--plugin zero2 \
--batch_size 8 \