v0.3.2

What's Changed

• ci: add manual coverage dispatch and fix broken coverage config by @voorhs in #325
• ci: fix Windows test jobs broken by bash-only shell step by @voorhs in #328
• test: cover JSONFormatter and macro retrieval metrics (Tier A) by @voorhs in #329
• test: cover augmentation CLIs + fix broken console-script entry points (Tier B) by @voorhs in #330
• test: cover Generator dump/load + async empty-response guard (Tier C) by @voorhs in #332
• fix(cache): handle directory cache entries (#326, #327) + tests (Tier D) by @voorhs in #331
• ci: gate combined coverage against an 85% regression floor (Tier E) by @voorhs in #333
• fix(cache): include model identity in structured-output cache key (#334) by @voorhs in #336
• fix(embedder): stabilize offline embedding cache key, prevent cross-model collision (#321) by @voorhs in #337
• release: prepare v0.3.2 by @voorhs in #338

Full Changelog: v0.3.1...v0.3.2

v0.3.1

What's Changed

See details at CHANGELOG.md.

• tmp fix for faiss typing by @voorhs in #292
• fix: update tunable decision test for optuna 4.9 by @voorhs in #293
• Fix/hf rate limit on tests by @voorhs in #294
• ci: bump GitHub Actions to Node 24 versions by @voorhs in #300
• test: mock live-API tests (no live OpenAI/OpenSearch in CI) by @voorhs in #301
• fix(llm-scorer): persist generator_config across dump/load roundtrip by @voorhs in #302
• fix(tests): restore pipeline interruption tests after sampler API removal by @voorhs in #303
• ci: enforce mypy --strict on tests/ (closes #296-class drift) by @voorhs in #316
• test: remove setup_environment helper, use pytest tmp_path by @voorhs in #317
• ci: gate full OS/Python matrix behind full-ci label and dev pushes by @voorhs in #318
• ci: type-check user_guides and docs under a soft mypy profile by @voorhs in #320
• release: prepare v0.3.1 by @voorhs in #319

Full Changelog: v0.3.0...v0.3.1

v0.3.0

What's Changed

See details at CHANGELOG.md.

• Update conf.py by @Samoed in #255
• Feat/mcp interface by @voorhs in #260
• Feat/open search by @voorhs in #259
• Refactor/migrate to uv by @voorhs in #257
• feat/train-embeddings by @k0lenk4 in #246
• Feat/gcn scorer by @SeBorgey in #261
• adversarial augmentation by @Tetragrammaton123 in #251
• Feat/api embeddings by @voorhs in #263
• update dependencies by @Samoed in #272
• Activate TC rules by @Samoed in #273
• F/split readiness utility by @voorhs in #276
• upd openai version by @voorhs in #277
• F/multilable split readiness check by @voorhs in #278
• add from_preset to OptimizationConfig by @voorhs in #279
• lazy import heavy dependencies by @voorhs in #267
• F/always separate oos by @voorhs in #280
• Add vllm support by @Samoed in #281
• F/OpenAI safe batching by @voorhs in #282
• B/fallback tokenizer for OpenAI embedder by @voorhs in #283
• Pin yarl below 1.24.0 until PyPI wheels are complete by @voorhs in #285
• Fix/sentence transformers embedder finetuning by @voorhs in #284
• docs: CHANGELOG and Sphinx release 0.3.0 by @voorhs in #286
• docs: installation extras guide by @voorhs in #287
• docs: quickstart, concepts, adversarial augmentation by @voorhs in #288
• docs: refresh embedder configuration user guide by @voorhs in #289
• docs: HTTP and MCP server deployment guide by @voorhs in #290

New Contributors

• @k0lenk4 made their first contribution in #246
• @Tetragrammaton123 made their first contribution in #251

Full Changelog: v0.2.0...v0.3.0

v0.2.0

New Scorers

• bert, lora, peft - by @voorhs, @SeBorgey, @nikiduki, @riapush
• rnn, cnn - by @voorhs, @SeBorgey
• catboost - by @Samoed, @nikiduki
• zero shot methods (bi encoder, cross encoder, llm) - by @voorhs, @Darinochka

AutoML

• optimization presets by @voorhs
• refactor HPO schema by @voorhs
• autointent interruption handling by @Samoed

Other

• node validation by @Samoed
• codecarbon callback by @Darinochka, @Samoed

Chores

• updated docs and tutorials by @voorhs
• optimized tests for repo by @Samoed
• innumerable bugs fixes by @voorhs, @Samoed

Full Changelog: v0.1.0...v0.2.0

v0.1.0

New functionality

• optuna samplers: TPE, Random, Brute
• cross-validation (previously: only hold-out validation)
• basic presets for balancing between the quickness and the quality
• logging to wandb and tensorboard
• LLM-based augmentation strategies for enriching your training data

Improvements

• better regular expressions support
• better UX on conducting experiments
• more convenient way to dump fitted pipeline to disk and then load it for inference

Documentation

Check out our updated user guides!

v0.0.1

Features

• Library of intent classification methods:
  • regexp module for rule-based classification
  • proxy tuning hyperparams of embedding model using retrieval metrics
  • scoring modules for predicting intents probabilities
  • decision-making modules for constructing final prediction for multi-class and multi-label classification and out-of-domain detection
• Auto ML approach to creating intent classifier:
  • greedy optimization for tuning hyperparameters
  • no target leakage, thanks to hold-out validation
  • embeddings caching for improving efficiency
  • both Python API and CLI
• Basic but flexible inference with automatically configured intent classifier
• Easy data manipulation with hugging face datasets integrated

Documentation

• API Reference for all modules, metrics and etc
• User guides with basic and advanced usage both for Python API and CLI
• Theoretical sections on dialogue systems creation and key concepts of AutoIntent