wehub-resource-sync
42 lines
1.2 KiB
Bash
42 lines
1.2 KiB
Bash
# 2 * 21GiB
|
|
PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
|
|
IMAGE_MAX_TOKEN_NUM=1024 \
|
|
VIDEO_MAX_TOKEN_NUM=128 \
|
|
FPS_MAX_FRAMES=16 \
|
|
NPROC_PER_NODE=2 \
|
|
CUDA_VISIBLE_DEVICES=0,1 \
|
|
swift sft \
|
|
--model Qwen/Qwen3-VL-4B-Instruct \
|
|
--dataset 'AI-ModelScope/alpaca-gpt4-data-zh#10000' \
|
|
'AI-ModelScope/LaTeX_OCR:human_handwrite#5000' \
|
|
'swift/VideoChatGPT:Generic#2000' \
|
|
--load_from_cache_file true \
|
|
--split_dataset_ratio 0.01 \
|
|
--tuner_type lora \
|
|
--torch_dtype bfloat16 \
|
|
--num_train_epochs 1 \
|
|
--per_device_train_batch_size 1 \
|
|
--per_device_eval_batch_size 1 \
|
|
--attn_impl flash_attn \
|
|
--padding_free true \
|
|
--learning_rate 1e-4 \
|
|
--lora_rank 8 \
|
|
--lora_alpha 32 \
|
|
--target_modules all-linear \
|
|
--freeze_vit true \
|
|
--freeze_aligner true \
|
|
--packing true \
|
|
--gradient_checkpointing true \
|
|
--vit_gradient_checkpointing false \
|
|
--gradient_accumulation_steps 2 \
|
|
--eval_steps 100 \
|
|
--save_steps 100 \
|
|
--save_total_limit 2 \
|
|
--logging_steps 5 \
|
|
--max_length 4096 \
|
|
--output_dir output \
|
|
--warmup_ratio 0.05 \
|
|
--deepspeed zero2 \
|
|
--dataset_num_proc 4 \
|
|
--dataloader_num_workers 4
|