6a391147a6
minor: refactoring big changes.
Build and Publish ray-serve-apps / build-and-publish (push) Successful in 12s
2026-02-12 18:47:50 -05:00
297b0d8ebd
fix: move mlflow import inside __init__ to avoid cloudpickle serialization failure
...
Build and Publish ray-serve-apps / build-and-publish (push) Successful in 16s
The strixhalo LLM worker uses py_executable which bypasses pip runtime_env.
Module-level try/except still fails because cloudpickle on the head node
resolves the real InferenceLogger class and serializes a module reference.
Moving the import inside __init__ means it runs at actor construction time
on the worker, where ImportError is caught gracefully.
2026-02-12 07:06:49 -05:00
15e4b8afa3
fix: make mlflow_logger import optional with no-op fallback
...
Build and Publish ray-serve-apps / build-and-publish (push) Successful in 11s
The strixhalo LLM worker uses py_executable pointing to the Docker
image venv which doesn't have the updated ray-serve-apps package.
Wrap all InferenceLogger imports in try/except and guard usage with
None checks so apps degrade gracefully without MLflow logging.
2026-02-12 07:01:17 -05:00
7ec2107e0c
feat: add MLflow inference logging to all Ray Serve apps
...
Build and Publish ray-serve-apps / build-and-publish (push) Successful in 16s
- Add mlflow_logger.py: lightweight REST-based MLflow logger (no mlflow dep)
- Instrument serve_llm.py with latency, token counts, tokens/sec metrics
- Instrument serve_embeddings.py with latency, batch_size, total_tokens
- Instrument serve_whisper.py with latency, audio_duration, realtime_factor
- Instrument serve_tts.py with latency, audio_duration, text_chars
- Instrument serve_reranker.py with latency, num_pairs, top_k
2026-02-12 06:14:30 -05:00
2edafc33c0
async vllm is better.
Build and Publish ray-serve-apps / build-and-publish (push) Successful in 1m3s
2026-02-11 06:05:50 -05:00
c9d7a2b5b7
fixing coqui
Build and Publish ray-serve-apps / build-and-publish (push) Failing after 20s
2026-02-09 09:14:30 -05:00
4549295a07
trigger: test package upload after gitea temp fix
Build and Publish ray-serve-apps / build-and-publish (push) Successful in 5m16s
2026-02-03 20:12:30 -05:00
665416bb0e
chore: trigger build with repo secrets
Build and Publish ray-serve-apps / build-and-publish (push) Failing after 50s
2026-02-03 19:33:45 -05:00
e853b805ae
chore: trigger pipeline with org-level runner
Build and Publish ray-serve-apps / build-and-publish (push) Failing after 2m50s
2026-02-03 19:22:34 -05:00
9bc40cfd20
chore: trigger rebuild after gitea storage migration
Build and Publish ray-serve-apps / build-and-publish (push) Failing after 46s
2026-02-03 16:07:27 -05:00
4a560f9b9e
chore: retrigger pipeline after runner restart
Build and Publish ray-serve-apps / build-and-publish (push) Failing after 12m51s
2026-02-03 15:49:43 -05:00
baf86e5609
ci: semver based on commit message keywords
...
Build and Publish ray-serve-apps / build-and-publish (push) Failing after 14m11s
- 'major' in message -> increment major, reset minor/patch
- 'minor' or 'feature' -> increment minor, reset patch
- 'bug', 'chore', anything else -> increment patch
- Release number from git rev-list commit count
- Format: major.minor.patch+release
2026-02-03 15:25:15 -05:00
3fb6d8f9c2
chore: trigger rebuild after S3 storage migration
2026-02-03 15:12:54 -05:00
8ef914ec12
feat: initial ray-serve-apps PyPI package
...
Build and Publish ray-serve-apps / lint (push) Failing after 11m2s
Build and Publish ray-serve-apps / publish (push) Has been cancelled
Implements ADR-0024: Ray Repository Structure
- Ray Serve deployments for GPU-shared AI inference
- Published as PyPI package for dynamic code loading
- Deployments: LLM, embeddings, reranker, whisper, TTS
- CI/CD workflow publishes to Gitea PyPI on push to main
Extracted from kuberay-images repo per ADR-0024
2026-02-03 07:03:39 -05:00
eac8f27f2e
Initial commit
2026-02-03 11:59:56 +00:00
