Releases: mostly-ai/mostlyai
Releases · mostly-ai/mostlyai
4.2.6
4.2.5
4.2.4
What's Changed
🚀 Features
- enable filtering of lists by owner id #264
- prompt users if api_key is missing during SDK initialization #270
Full Changelog: 4.2.3...4.2.4
4.2.3
4.2.2
4.2.1
What's Changed
🐛 Fixes
- updated UDS path to avoid AF_UNIX path too long error #246
- fixed issue with seed for language-only setups #245
- fixed conversion of Spark DF in GeneratorConfig to base64 string #247
- simplified rebalancing column validation in synthetic table config #249
- fixed execution plan of probe tasks #255
📦 Uncategorized
- Update README.md #254
Full Changelog: 4.2.0...4.2.1
4.2.0
What's Changed
🚀 Features
- support new non-TEXT encoding types for LANGAUGE models #233
🐛 Fixes
- support Spark DF from a Spark Connect session #242
Full Changelog: 4.1.7...4.2.0
4.1.7
What's Changed
🚀 Features
- handle Spark DF (also in serverless mode)
- support seed or size argument in combination with config #234
🐛 Fixes
- avoid redundant df to base64 conversion #232
📦 Uncategorized
- change definition of
ExecutionPlan
/Task
#192
Full Changelog: 4.1.3...4.1.7
4.1.3
What's Changed
- feat: fix model configurations instead of raising exception
- feat: consider only included columns on SourceTableConfig
- feat: add g.training.logs() and sd.generation.logs() for downloading log files
- feat: add display option to .reports()
- build: add
ipykernel
todev
deps andipywidgets
to general deps - chore: sync with latest Public API (limit column and table names to 256 chars)
- chore: use same python executable for CLI calls
- docs: prune deprecated local_port arg from docs
- docs: refine pip instructions for tutorials
- docs: add star schema correlations tutorial
Full Changelog: 4.1.1...4.1.3