phoenix-oss/llama-stack
1
0
Fork 0
forked from phoenix-oss/llama-stack-mirror
Code Pull requests Releases Packages 2 Activity Actions

Use inference APIs for executing Llama Guard (#121)

Browse source 
We should use Inference APIs to execute Llama Guard instead of directly needing to use HuggingFace modeling related code. The actual inference consideration is handled by Inference.
This commit is contained in:
Ashwin Bharambe 2024-09-28 15:40:06 -07:00 • committed by GitHub
parent 6236634d84
commit 0a3999a9a4
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
9 changed files with 167 additions and 204 deletions
Download patch file Download diff file Expand all files Collapse all files

4
llama_stack/providers/impls/meta_reference/safety/safety.py
View file

@ -88,10 +88,10 @@ class MetaReferenceSafetyImpl(Safety):
assert (
cfg is not None
), "Cannot use LlamaGuardShield since not present in config"
model_dir = resolve_and_get_path(cfg.model)
return LlamaGuardShield(
model_dir=model_dir,
model=cfg.model,
inference_api=self.inference_api,
excluded_categories=cfg.excluded_categories,
disable_input_check=cfg.disable_input_check,
disable_output_check=cfg.disable_output_check,