phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-08-11 04:28:02 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

torchtune: save job status on schedule

Browse source 
Otherwise a failed job never gets registered to extract status, even if
it's still present in job_list.

Signed-off-by: Ihar Hrachyshka <ihar.hrachyshka@gmail.com>
This commit is contained in:
Ihar Hrachyshka 2025-02-18 09:26:56 -05:00
parent 80070b966a
commit 07a1d44f4c
1 changed files with 1 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

3
llama_stack/providers/inline/post_training/torchtune/post_training.py
View file

@ -65,6 +65,7 @@ class TorchtunePostTrainingImpl:
status=JobStatus.scheduled, status=JobStatus.scheduled,
scheduled_at=datetime.now(), scheduled_at=datetime.now(),
) )
self.jobs_status[job_uuid] = job_status_response
self.jobs_list.append(post_training_job) self.jobs_list.append(post_training_job)
if isinstance(algorithm_config, LoraFinetuningConfig): if isinstance(algorithm_config, LoraFinetuningConfig):
@ -100,8 +101,6 @@ class TorchtunePostTrainingImpl:
else: else:
raise NotImplementedError() raise NotImplementedError()
self.jobs_status[job_uuid] = job_status_response
return post_training_job return post_training_job
async def preference_optimize( async def preference_optimize(