FMInference / FlexLLMGen

Running large language models on a single GPU for throughput-oriented scenarios.
Apache License 2.0
9.21k stars 547 forks source link

replace Argparse by Fire #21

Closed Borda closed 1 year ago

Borda commented 1 year ago

Simplify the CLI with Fire which has zero other dependencies; also it pulls typing, default and docsstrings from the function footprint so reducing the duplicity :chipmunk:

Borda commented 1 year ago

@Ying1123 @merrymercy can I get your openion/review? :otter:

Ying1123 commented 1 year ago

Hi Borda, thanks for the PR. I'm sorry that I need to temporarily revert it, the following command is broken up. python3 chatbot.py --model facebook/opt-30b --percent 0 100 100 0 100 0