File size: 382 Bytes
cb2428f |
1 2 3 4 5 6 7 8 9 10 11 12 13 |
CUDA_VISIBLE_DEVICES=0 \
swift export \
--model Qwen/Qwen2.5-72B-Instruct \
--dataset 'AI-ModelScope/alpaca-gpt4-data-zh#500' \
'AI-ModelScope/alpaca-gpt4-data-en#500' \
--device_map cpu \
--quant_n_samples 256 \
--quant_batch_size 1 \
--max_length 2048 \
--quant_method awq \
--quant_bits 4 \
--output_dir Qwen2.5-72B-Instruct-AWQ
|