This commit is contained in:
@@ -0,0 +1,14 @@
|
||||
pip install "transformers<4.52"
|
||||
|
||||
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
|
||||
Reference in New Issue
Block a user