phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-07-30 23:51:00 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix evals and scoring

Browse source
This commit is contained in:
Dinesh Yeduguru 2024-11-12 18:14:58 -08:00
parent 55d66ca918
commit 606df220f5
3 changed files with 2 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

2
llama_stack/providers/inline/eval/meta_reference/eval.py
View file

@ -150,7 +150,7 @@ class MetaReferenceEvalImpl(Eval, EvalTasksProtocolPrivate):
messages.append(candidate.system_message)
messages += input_messages
response = await self.inference_api.chat_completion(
model=candidate.model,
model_id=candidate.model,
messages=messages,
sampling_params=candidate.sampling_params,
)

2
llama_stack/providers/inline/scoring/llm_as_judge/scoring_fn/llm_as_judge_scoring_fn.py
View file

@ -62,7 +62,7 @@ class LlmAsJudgeScoringFn(BaseScoringFn):
)
judge_response = await self.inference_api.chat_completion(
model=fn_def.params.judge_model,
model_id=fn_def.params.judge_model,
messages=[
{
"role": "user",

1
llama_stack/providers/remote/inference/vllm/vllm.py
View file

@ -85,7 +85,6 @@ class VLLMInferenceAdapter(Inference, ModelRegistryHelper, ModelsProtocolPrivate
logprobs: Optional[LogProbConfig] = None,
) -> AsyncGenerator:
model = await self.model_store.get_model(model_id)
print(f"model={model}")
request = ChatCompletionRequest(
model=model.provider_resource_id,
messages=messages,