wehub-resource-sync
30 lines
859 B
Bash
30 lines
859 B
Bash
CUDA_VISIBLE_DEVICES=0 \
|
|
swift sft \
|
|
--model deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B \
|
|
--tuner_type full \
|
|
--dataset AI-ModelScope/function-calling-chatml \
|
|
--load_from_cache_file true \
|
|
--split_dataset_ratio 0.01 \
|
|
--agent_template react_en \
|
|
--loss_scale react \
|
|
--response_prefix '' \
|
|
--torch_dtype bfloat16 \
|
|
--num_train_epochs 2 \
|
|
--per_device_train_batch_size 1 \
|
|
--per_device_eval_batch_size 1 \
|
|
--learning_rate 1e-5 \
|
|
--gradient_accumulation_steps 8 \
|
|
--eval_steps 100 \
|
|
--save_steps 100 \
|
|
--save_total_limit 2 \
|
|
--logging_steps 5 \
|
|
--max_length 8192 \
|
|
--save_only_model true \
|
|
--packing true \
|
|
--use_liger_kernel true \
|
|
--output_dir output \
|
|
--warmup_ratio 0.05 \
|
|
--attn_impl flash_attn \
|
|
--dataloader_num_workers 4 \
|
|
--dataset_num_proc 16
|