phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-06-28 02:53:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix: Add type field to DPOAlignmentConfig for remote provider serialization

Browse source 
Addresses review feedback to keep preference optimization APIs DPO-specific
rather than accepting all algorithm types.
This commit is contained in:
Nehanth 2025-06-25 16:40:37 -04:00
parent 40a0cef38e
commit 243fe37a67
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
llama_stack/apis/post_training/post_training.py
View file

@ -125,7 +125,7 @@ class PostTrainingRLHFRequest(BaseModel):
validation_dataset_id: str validation_dataset_id: str
algorithm: RLHFAlgorithm algorithm: RLHFAlgorithm
algorithm_config: AlgorithmConfig algorithm_config: DPOAlignmentConfig
optimizer_config: OptimizerConfig optimizer_config: OptimizerConfig
training_config: TrainingConfig training_config: TrainingConfig
@ -202,7 +202,7 @@ class PostTraining(Protocol):
self, self,
job_uuid: str, job_uuid: str,
finetuned_model: str, finetuned_model: str,
algorithm_config: AlgorithmConfig, algorithm_config: DPOAlignmentConfig,
training_config: TrainingConfig, training_config: TrainingConfig,
hyperparam_search_config: dict[str, Any], hyperparam_search_config: dict[str, Any],
logger_config: dict[str, Any], logger_config: dict[str, Any],