cloudstack-llc/mlx-knife
1
0
Fork 0
mirror of https://github.com/cloudstack-llc/mlx-knife.git synced 2026-08-27 02:01:29 -04:00
Code Issues Packages Projects Releases Wiki Activity
64 Commits 4 Branches 53 Tags
dev
Commit Graph

12 Commits

Author SHA1 Message Date
The BROKE Cluster Team 7f10187bee fix: Runtime gates + unit tests; benchmark GPU analysis 
Core:
- Run preflight passes probe/framework to audio_runtime_compatibility
- STT model_type gate extended (vibevoice, audio)
- MLX 0.30.x compat: catch Exception in whisper_tokenizer
- Embedding-gate unit tests (3 tests)
- Removed get_encoding duplication (-45 LOC)

Benchmark:
- GPU analysis section in reports
 2026-02-07 23:38:34 +01:00
The BROKE Cluster Team e021fb32cd Release 2.0.4-beta.10: Audio PyPI fix (tiktoken workaround complete) 
Audio/Whisper works with pip install - no Git workaround needed.
See CHANGELOG.md for details.

Tested: 647 passed, 11 skipped (Python 3.10-3.12)
 2026-02-05 10:42:50 +01:00
The BROKE Cluster Team bf7480d042 Release 2.0.4-beta.9: Audio transcription via mlx-audio 
Major Features:
- Audio transcription via mlx-audio backend (Whisper, >10min duration)
- OpenAI /v1/audio/transcriptions endpoint
- Memory Gate System (Vision: 8GB, Audio: 4GB)
- Config-based backend routing (ADR-020)
- Benchmark toolchain (memmon/memplot, Schema v0.2.2)

Key Fixes:
- EuroLLM tokenizer decoding
- Vision-model text-only routing regression
- Multimodal model context length detection
- Memory cleanup bug (mx.metal.clear_cache)
- Orphan process bug

Test Results:
- Unit tests: 647 passed, 11 skipped (Python 3.10-3.12)
- wet-umbrella: 171 passed total

See CHANGELOG.md for complete details and known issues.
 2026-02-04 03:10:30 +01:00
The BROKE Cluster Team e8b10ea10b Release 2.0.4-beta.8: Audio transcription support (experimental) 
Audio input via --audio flag (CLI) and input_audio content type (Server API).
Uses mlx-vlm native audio processing. ~30s duration limit (model constraint).
Currently only Gemma-3n tested (requires --repair-index fix).

Also includes:
- SERVER-HANDBOOK compliance (image limits, validation error envelopes)
- Dependency updates: mlx>=0.30.0, mlx-lm>=0.30.0, huggingface-hub>=1.0.0
- Audio E2E test suite + ADR-019
 2026-01-23 20:20:59 +01:00
The BROKE Cluster Team 5751545b8b Release 2.0.4-beta.7: Server robustness + Vision per-chunk streaming 
- Server: exit codes, /v1/models crash fix, vision routing, MLXK2_MAX_TOKENS
- Vision: true SSE streaming, hallucination fix (local numbering)
- Workspace: list prefix-match, push ambiguous pattern handling
- Docs: SERVER-HANDBOOK accuracy updates

See CHANGELOG.md for details.
 2026-01-18 16:57:32 +01:00
The BROKE Cluster Team 53d9cca82d Release 2.0.4-beta.6: Local workspace workflow + Vision batch processing 
- Complete local development cycle: clone → repair → run/show/server on
  workspace paths without HuggingFace round-trips
- Vision processing now defaults to safe chunking (one image at a time,
  prevents OOM + hallucination)
- Resumable clone with --force-resume and deterministic temp cache naming
- Improved test infrastructure (umbrella marker convention)
- 161 Wet Umbrella tests passing including new Vision→Geo pipe integration tests

See CHANGELOG.md for complete details.
 2026-01-07 17:11:07 +01:00
The BROKE Cluster Team 86f669dc82 Release 2.0.4-beta.1: Vision + Pipes + Memory 
- Vision Support (Issue #45): CLI + Server with OpenAI-compatible image API, EXIF metadata
- Unix Pipes (ADR-014): stdin support, isatty detection, SIGPIPE handling
- Memory-Aware Loading (ADR-016): Pre-load checks with >70% RAM warnings
- Python 3.9-3.14: Full compatibility verified (476-485 tests passing)
- Fixed: --log-json regression (Issue #44), Vision multimodal history filtering

See CHANGELOG.md for complete details.
 2025-12-16 19:35:30 +01:00
The BROKE Cluster Team 05f1c30486 Release 2.0.3: Foundation for pipes 
Foundation release for Unix pipe integration with stderr separation,
benchmark infrastructure, and reasoning control improvements.

Breaking Changes:
- stdout/stderr separation (Issue #43) - errors to stderr in human mode
- JSON mode unchanged (all output to stdout)

Features:
- Benchmark reporting infrastructure (ADR-013 Phase 0)
- --no-reasoning flag (Issue #40 partial - GPT-OSS/QwQ only)
- Interactive mode reasoning control (review_report.md fixes)

Bug Fixes:
- huggingface-hub 1.x incompatibility (critical dependency fix)
- Streaming parity tests refactored (Portfolio Discovery)

Testing:
- 308 tests passing (Python 3.9-3.13)
- 35 skipped (opt-in live tests)
- 79/91 E2E tests passing with HF_HOME

See CHANGELOG.md for complete details and migration guide.
 2025-11-17 22:54:06 +01:00
The BROKE Cluster Team 21cf188fcc Release 2.0.1: Portfolio Discovery + CLI Exit Code Fixes 
Issue #32: Stop token Portfolio Discovery validates generic fix across all models
- Auto-discovers MLX chat models in HF_HOME with 4-filter validation
- RAM-aware testing (40-70% budgets) prevents OOM
- Empirical report generation (stop_token_config_report.json)
- Fallback to 3 predefined models without HF_HOME
- Implementation: tests_2.0/test_stop_tokens_live.py (~110 LOC)

Issue #38: CLI exit codes now propagate run command errors correctly
- Both text and JSON modes return exit code 1 on model execution failures
- Fixed: run_model() now returns error strings in both modes
- Implementation: mlxk2/operations/run.py + mlxk2/cli.py error detection
- New tests: tests_2.0/test_cli_run_exit_codes.py (9 comprehensive tests)

Testing: 306 passed, 20 skipped (zero regressions)
Docs: Updated README, TESTING, SECURITY for 2.0.1 stable release
Version: 2.0.0 → 2.0.1 (mlxk2/__init__.py)
 2025-11-08 20:28:54 +01:00
The BROKE Cluster Team fb54f59cd4 Release 2.0.0-beta.6: Stop token & compatibility bug fixes 
Fixes Issue #32 (generic multi-EOS detection) and Issue #37 (model detection)

  - Generic stop token detection: Multi-EOS models (MXFP4, Qwen, Llama) now use eos_token_ids Set instead of
  model-specific workarounds
  - Private/org MLX model detection: `mlxk run` now works outside `mlx-community/*` namespace
  - Commit-pinned compatibility checks: Models with `@commit_hash` validated before inference
  - Packaging dependencies: Fixed `pip install -e .` requirements

  - ADR-009: Stop Token Detection Fix (generic approach + test strategy)
  - ADR-011: E2E Live Test Architecture (planned)

See CHANGELOG.md and TESTING.md for details.
 2025-10-24 15:46:42 +02:00
The BROKE Cluster Team f5fe1dd061 Release 2.0.0-beta.5: Enhanced error handling & bug fixes 
Features:
- Enhanced error handling & logging (ADR-004): Unified error envelope, structured logging with JSON support, request correlation
- Legacy format detection (Issue #37): Runtime compatibility check for weight file formats

Bug Fixes:
- Issue #37: Models with legacy weight formats now correctly detected as runtime-incompatible
- CLI regression fix: mlxk2 without arguments shows help instead of JSON error

Test Status: 295/295 passed, 14 skipped
 2025-10-21 00:24:47 +02:00
The BROKE Cluster Team 57bf6d86be 2.0.0-beta.3: Feature Complete - Full 1.1.1 Parity Achieved 
Major Features Added:
  • Complete run command implementation with interactive/single-shot modes
  • MLXRunner core engine ported from 1.x with modular architecture
  • OpenAI-compatible server with SIGINT-robust supervisor mode
  • Experimental push feature properly isolated behind environment variable

  Key Improvements:
  - Full feature parity with 1.1.1 stable releases
  - Enhanced human output formatting across all commands
  - Clean separation of stable (184 tests) vs experimental features
  - Updated demo GIF showcasing improved 2.0 interface

  Fixes:
  - Pull operation cache pollution (Issue #30) with preflight access checks
  - Test stability improvements across all environments

  Architecture:
  - Modular runner design with focused helper modules
  - Thread-safe model loading and memory management
  - stable testing across Python 3.9-3.13

  Ready for use as comprehensive 1.x alternative.
 2025-09-14 18:04:18 +02:00