Spaces:
Paused
Paused
File size: 1,033 Bytes
4721aa1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 |
#! /usr/bin/env bash
set -ex
PRE_SEQ_LEN=128
LR=2e-2
NUM_GPUS=1
MAX_SOURCE_LEN=1024
MAX_TARGET_LEN=128
DEV_BATCH_SIZE=1
GRAD_ACCUMULARION_STEPS=32
MAX_STEP=1000
SAVE_INTERVAL=500
DATESTR=`date +%Y%m%d-%H%M%S`
RUN_NAME=advertise_gen_pt
BASE_MODEL_PATH=THUDM/chatglm3-6b
DATASET_PATH=formatted_data/advertise_gen.jsonl
OUTPUT_DIR=output/${RUN_NAME}-${DATESTR}-${PRE_SEQ_LEN}-${LR}
mkdir -p $OUTPUT_DIR
torchrun --standalone --nnodes=1 --nproc_per_node=$NUM_GPUS finetune.py \
--train_format input-output \
--train_file $DATASET_PATH \
--preprocessing_num_workers 1 \
--model_name_or_path $BASE_MODEL_PATH \
--output_dir $OUTPUT_DIR \
--max_source_length $MAX_SOURCE_LEN \
--max_target_length $MAX_TARGET_LEN \
--per_device_train_batch_size $DEV_BATCH_SIZE \
--gradient_accumulation_steps $GRAD_ACCUMULARION_STEPS \
--max_steps $MAX_STEP \
--logging_steps 1 \
--save_steps $SAVE_INTERVAL \
--learning_rate $LR \
--pre_seq_len $PRE_SEQ_LEN 2>&1 | tee ${OUTPUT_DIR}/train.log
|