- `soup ui` command launches local web interface at http://127.0.0.1:7860
- FastAPI backend with REST API: runs, metrics, config validation, training
control, data inspection, templates, system info
- Self-contained SPA frontend (Dashboard, New Training, Data Explorer,
Model Chat) with Chart.js loss/LR charts
- Auto-opens browser on launch (--no-browser to disable)
- Config validation via new load_config_from_string() in config/loader.py
- 40 new tests (530 total), ruff clean
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add `modality: vision` config option for vision-language model training
- Add LLaVA and ShareGPT4V data format detection and conversion
- Add `image_dir` field in DataConfig for resolving image paths
- Add vision model loading via AutoModelForVision2Seq + AutoProcessor in SFT trainer
- Add `soup init --template vision` with LLaMA-3.2-Vision config
- Add image statistics display in `soup data inspect` for vision datasets
- Add Pillow as optional `vision` extra dependency
- Add Pillow to `soup doctor` dependency checks
- 51 new tests (455 total), ruff clean
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
`str | None` union syntax requires Python 3.10+. CI runs on 3.9.
Added `from __future__ import annotations` to fix.
Also added Release Checklist section to CLAUDE.md.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add --early-stop flag to soup sweep: skip remaining runs when loss exceeds best by threshold
- Add Rich progress bars for HuggingFace Hub model downloads in SFT/DPO trainers
- Pin click>=8.2.0 to fix typer compatibility (11 tests were failing)
- Bump version 0.3.2 → 0.4.0
- 329 tests passing, lint clean
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
pytest junitxml wraps results in <testsuites><testsuite tests="N">,
the count is on the child element, not the root.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- `soup version --full` shows version, Python, GPU backend, installed extras
- Dynamic test count badge via Gist endpoint in CI
- README: Optional Extras table, --verbose note, CSV/Parquet, Changelog link
- 323 tests passing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
CI now parses pytest junitxml output and updates a public Gist with the
exact test count. Shields.io reads the Gist as an endpoint badge.
Requires GIST_TOKEN secret (PAT with gist scope).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Replace hardcoded "321 passed" badge with "passing"
- Add Optional Extras table (serve, data, eval, deepspeed, dev)
- Add note that --verbose must go before command name
- Mention CSV/Parquet support in Data Formats
- Add Changelog section linking to GitHub Releases
- Remove duplicate commands from All Commands (resume, wandb, deepspeed variants)
- Remove runs delete from commands list (rarely used)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Rich markup splits --verbose with escape codes on CI, check for
"verbose" instead of "--verbose".
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Friendly error messages: wrap all commands in try/except, map known errors
(CUDA OOM, missing deps, connection errors) to 2-3 line messages with fix hints
- Global --verbose flag for full tracebacks
- soup doctor: check system info, GPU, all dependency versions with fix suggestions
- soup quickstart: one-command demo (creates data + config + trains TinyLlama)
- Confirmation prompts before train/sweep (skip with --yes)
- 40 new tests (321 total), all passing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- soup serve: FastAPI inference server with OpenAI-compatible API, SSE streaming
- soup data generate: synthetic data generation via OpenAI API or local models
- soup sweep: grid/random hyperparameter search with experiment tracker integration
- soup diff: side-by-side model comparison with metrics
- Multi-GPU/DeepSpeed: ZeRO Stage 2/3 configs, --deepspeed flag in train command
- 95 new tests (281 total), all passing
- Removed TESTING_GUIDE.md
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Replace torch_dtype with dtype in chat, merge, export commands (transformers deprecation)
- Replace warmup_ratio with warmup_steps in SFT/DPO trainers (transformers v5.2 deprecation)
- Add TESTING_GUIDE.md for local Windows testing workflow
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
PyTorch 2.7+ renamed `get_device_properties().total_mem` to
`total_memory`. Fixed in gpu.py and callback.py.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Rich markup splits --resume/--wandb with escape codes in CI (Python 3.9).
Use case-insensitive substring match instead of exact flag match.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Triggers on `git tag v0.2.0 && git push --tags`.
Uses PyPI trusted publisher (OIDC) — no token needed in secrets.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Change primary install to `pip install soup-cli`
- Move GitHub install to secondary option
- Add PyPI version badge
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add Python/License/Tests/CI badges
- Add "All Commands" section with full CLI reference
- Add lint command to Development section
- Update test count to 147
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The 4 hex char suffix (65536 possibilities) caused a collision when
generating 100 IDs within the same second on fast CI runners.
Increased to 8 hex chars (4 billion possibilities).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Fix 23 ruff errors: line too long, unused imports, ambiguous vars
- Fix validator: empty string is valid data, only count None as empty
- Remove unused imports in display.py and validator.py
- Rename ambiguous `l` vars to `part`, `entry`, `length`
- Break long lines in callback.py, display.py, sft.py, constants.py
All 20 tests passing, ruff clean.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Typer CLI: soup init, soup train, soup data inspect/validate
- Pydantic config schema with YAML loader and validation
- Data pipeline: JSONL/JSON/CSV/Parquet + HuggingFace datasets
- Format detection: Alpaca, ShareGPT, ChatML (auto-detect)
- SFT trainer wrapper over transformers + peft + trl
- QLoRA/LoRA support with auto batch size estimation
- GPU detection (CUDA/MPS/CPU) and memory calculation
- Rich live terminal dashboard for training monitoring
- Config templates: chat, code, medical
- Tests (pytest) + GitHub Actions CI
- MIT license
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Alpamys