Qwen2.5-0.5B-Instruct-CTX-Int8 / run_qwen2.5_0.5b_gptq_int8_ctx_ax650.sh
qqc1989's picture
Upload 36 files
d8727b9 verified
./main_ax650 \
--system_prompt "You are Qwen, created by Alibaba Cloud. You are a helpful assistant." \
--template_filename_axmodel "qwen2.5-0.5b-gptq-int8-ctx-ax650/qwen2_p128_l%d_together.axmodel" \
--axmodel_num 24 \
--tokenizer_type 2 \
--url_tokenizer_model "http://127.0.0.1:12345" \
--filename_post_axmodel "qwen2.5-0.5b-gptq-int8-ctx-ax650/qwen2_post.axmodel" \
--filename_tokens_embed "qwen2.5-0.5b-gptq-int8-ctx-ax650/model.embed_tokens.weight.bfloat16.bin" \
--tokens_embed_num 151936 \
--tokens_embed_size 896 \
--use_mmap_load_embed 0 \
--live_print 1
# --kvcache_path /home/axera/ax-llm/build/kvcache_yuanqi \