What would you like to be added:
Config.md docs could have a more clear definition of the shared_prefix datagen along with a practical example.
The good example that others should change to use:
|
num_groups: 10 # Number of distinct shared prefixes |
Why is this needed: Clear docs
What would you like to be added:
Config.md docs could have a more clear definition of the shared_prefix datagen along with a practical example.
The good example that others should change to use:
inference-perf/examples/vllm/config-shared-prefix.yml
Line 21 in dc512a1
Why is this needed: Clear docs