wehub-resource-sync
26 lines
720 B
Bash
26 lines
720 B
Bash
CUDA_VISIBLE_DEVICES=0 \
|
|
MAX_NEGATIVE_SAMPLES=1 \
|
|
MAX_PIXELS=602112 \
|
|
swift sft \
|
|
--model Qwen/Qwen2.5-VL-3B-Instruct \
|
|
--task_type reranker \
|
|
--loss_type pointwise_reranker \
|
|
--tuner_type lora \
|
|
--dataset swift/TextCaps:rerank \
|
|
--split_dataset_ratio 0.05 \
|
|
--eval_strategy steps \
|
|
--output_dir output \
|
|
--eval_steps 100 \
|
|
--num_train_epochs 1 \
|
|
--save_steps 200 \
|
|
--per_device_train_batch_size 8 \
|
|
--per_device_eval_batch_size 8 \
|
|
--gradient_accumulation_steps 4 \
|
|
--padding_free true \
|
|
--attn_impl flash_attn \
|
|
--dataset_num_proc 8 \
|
|
--learning_rate 6e-5 \
|
|
--label_names labels \
|
|
--dataloader_drop_last true \
|
|
--attn_impl flash_attn
|