v1.2.4

What's Changed

• docs: rename Evaluation Suite to LLM Judge Suite; reorder + retitle by @liana313 in #265
• Bump version to 1.2.4 by @liana313 in #266

Full Changelog: v1.2.3...v1.2.4

v1.2.3

What's Changed

• Agentic map-reduce (Corpus + tools/REPL), docs, and README revamp by @liana313 in #264

Full Changelog: v1.2.2...v1.2.3

v1.2.2

What's Changed

• ci: pin dependencies to uv.lock for deterministic CI by @liana313 in #259
• Fix gpt-5/reasoning-model accuracy: model-aware max_tokens default by @liana313 in #260
• Experiment scripts for Biodex and Reranking by @StanChan03 in #219
• Consolidate duplicate benchmark directories (benchmark/ -> benchmarks/) by @liana313 in #261
• Fix flaky test_pairwise_judge (ambiguous row removed) by @liana313 in #262
• Bump version to 1.2.2 by @liana313 in #263

Full Changelog: v1.2.1...v1.2.2

v1.2.1

What's Changed

• Fix pairwise_judge GEPA helper-instruction tests (A/B placeholders) by @liana313 in #257
• Bump version to 1.2.1 by @liana313 in #258

Full Changelog: v1.2.0...v1.2.1

v1.2.0

What's Changed

• inconsistent sem_extract output fix by @StanChan03 in #206
• migration to uv from pip by @StanChan03 in #214
• fix typing in postprocessor by @StanChan03 in #222
• update dependency for colab env by @StanChan03 in #225
• add max_completion_token param for gpt5 models by @StanChan03 in #226
• manual cost calculations by @StanChan03 in #216
• add add for lmstats by @harshitgupta412 in #236
• Fix incorrect assertion in lm.get_completion by @harshitgupta412 in #235
• add start and end date param for web search by @harshitgupta412 in #239
• Implement long context handling in sem_agg by @harshitgupta412 in #238
• Feature/tpm rate limiting by @harshitgupta412 in #240
• add pubmed search api support by @ankitasun in #242
• add web extract function by @harshitgupta412 in #244
• Add AST-based lazy evaluation and semantic operator optimizations by @harshitgupta412 in #246
• Fix pairwise_judge to use A/B instead of col_A/col_B. col_A/col_B use 2 tokens, so cascade fails by @harshitgupta412 in #248
• Add scoped settings context manager with concurrency safety (#247) by @liana313 in #249
• fix pairwise judge benchmark to use A/B instead of True/False by @harshitgupta412 in #251
• Chore/docs examples by @harshitgupta412 in #250
• Fix mypy errors from dependency drift (unblock CI) by @liana313 in #254
• Bump version to 1.2.0 by @liana313 in #252
• Add release-on-tag PyPI publish workflow by @liana313 in #253
• Fix PyPI publish: use gepa from PyPI instead of git URL by @liana313 in #256

Full Changelog: v1.1.4...v1.2.0

v1.1.4

What's Changed

• Delete .idea folder, it was merged previously by mistake. by @vincentzed in #175
• Add explicit rate limiting functionality to LM class by @ankitasun in #185
• Fix logprobs bug from litellm version upgrade by @sidjha1 in #188
• [WIP] Support for Document Chunking by @StanChan03 in #190
• [WIP] Code clean up by @liana313 in #192
• Support zero-shot CoT for filter by @sidjha1 in #197
• eval suite by @harshitgupta412 in #207
• fix timeout by @harshitgupta412 in #208

New Contributors

• @vincentzed made their first contribution in #175

Full Changelog: v1.1.3...v1.1.4

v1.1.3

What's Changed

• Add example of Local Weaviate VS by @sidjha1 in #152
• Run ruff format. by @vincent-4 in #154
• added support for You.com, Bing and Tavily web search engines by @ankitasun in #149
• Correct return keys for tavily by @harshitgupta412 in #162
• Add Documentation on LLM Usage Limits by @sidjha1 in #161
• Added dependencies for web search engines by @ankitasun in #163
• DeepSeek CoT Support for Filter and Map by @StanChan03 in #142
• add tests for group by in sem_topk and sem_Agg by @harshitgupta412 in #170
• top_k deepseek cot support by @StanChan03 in #171
• Add Qdrant VS by @sidjha1 in #153
• Updated CONTRIBUTING.md to reflect now git org. by @dmatch01 in #178
• Update requirements.txt upgrading litellm version number to fix litellm provider bugs for WatsonX by @dmatch01 in #181

New Contributors

• @vincent-4 made their first contribution in #154
• @ankitasun made their first contribution in #149
• @dmatch01 made their first contribution in #178

Full Changelog: v1.1.0...v1.1.3

v1.1.0

Stable release