2 Commits

Author	SHA1	Message	Date
Alpamys	15a6daf342	feat: v0.14.0 — pre-training + MoE support ... Add continued pre-training task and Mixture of Experts model support: - `task: pretrain` for continued pre-training on raw text data - `plaintext` data format ({"text": "..."} JSONL or .txt files) - MoE model detection (Mixtral, Qwen3 MoE, DeepSeek V3, DBRX, OLMoE) - ScatterMoE LoRA (`moe_lora: true`) targets expert FFN + attention layers - `moe_aux_loss_coeff` for router load-balancing loss - Templates: `soup init --template pretrain` and `--template moe` - 85 new tests across test_pretrain.py and test_moe.py (1002 total) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-25 22:26:01 +05:00
Alpamys	5f728c04cc	Expand test suite from ~70 to 147 tests, fix flaky ordering bug ... Add 8 new test files covering previously untested modules: - test_chat.py: _detect_base_model, adapter validation - test_push.py: _format_size, _generate_model_card, token checks - test_init.py: all templates, overwrite confirm/deny, YAML validation - test_callback.py: SoupTrainerCallback with mocks - test_display.py: TrainingDisplay rendering + edge cases - test_loader.py: JSON/CSV/JSONL loading, empty lines, bad JSON - test_validator.py: validate_and_stats, extended_stats, _percentile - test_formats.py: reverse conversion, round-trips, edge cases Fix flaky test_list_runs_ordering by adding rowid DESC as tiebreaker in list_runs SQL query (runs created in same second had nondeterministic order). Update CLAUDE.md with test file inventory. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-02 20:52:33 +05:00