v0.3.9

• PR #1367: v0.3.9 - new release
• PR #1368: Migrate ChatGPT function to openai v1.0
• PR #1361: Adding changes for Flaml Sklearn integration
• PR #1258: Add feedback for forecasting
• PR #1364: Fix python3.8 failing testcases due to type hint
• PR #1308: Job scheduler implementation
• PR #1352: Make Logic Operators Case Insensitve
• PR #1342: Add train scores for ludwig in the create function handler.
• PR #1067: Verified that issue #1067 is resolved and added documentation for load pdf functionality.
• PR #1334: logging an error message for invalid files while loading
• PR #1330: [WIP] Improving error handling messages for Custom Functions
• PR #1033: feat: third party app support in EVADB
• PR #1339: Support semicolon and escaped strings in lark
• PR #1337: CREATE INDEX IF NOT EXISTS is broken.
• PR #1333: Fix #1333 dependency and CMD in DockerFile
• PR #1336: Add test_eva_db to gitignore
• PR #1327: Add the validation score and training time for create_function in XGBoost
• PR #1322: String Helper functions in EvaDB
• PR #1321: REST API Documentation
• PR #1323: Added docs for SET and SHOW CONFIG query
• PR #1316: Adding colab notebook for XGBoost Regression and Classification
• PR #1269: Add Milvus integration for vector create and search
• PR #1319: refactor: code refactor
• PR #1298: Convert nested join in Vector Queries to Pandas Merge.
• PR #1305: Starting the changes for XGBoost classification integration.
• PR #1289: SnowFlake Integration for EvaDB
• PR #1307: Fix: minor typo
• PR #1272: Configuration Manager Redesign
• PR #1283: Fix current issues with forecasting
• PR #1304: Bump Version to v0.3.9+dev
• PR #1303: v0.3.8 - new release

v0.3.8

• PR #1303: v0.3.8 - new release
• PR #1302: Reenable batch for release
• PR #1301: Add Documentation for UDF Unit Testing and Mocking
• PR #1232: Starting the change for XGBoost integration into EVADb.
• PR #1294: fix: improve testcase
• PR #1293: fix: make the table/function catalog insert operation atomic
• PR #1295: feat: add support for show databases
• PR #1296: feat: function_metadata supports boolean and float
• PR #1290: fix: text_summarization uses drop udf
• PR #1240: Add stable diffusion integration
• PR #1285: Update custom-ai-function.rst
• PR #1234: Added basic functionalities of REST apis
• PR #1281: Clickhouse integration
• PR #1273: Update custom-ai-function.rst
• PR #1274: Fix Notebook and Ray testcases at staging
• PR #1264: SHOW command for retrieveing configurations
• PR #1270: fix: Catalog init introduces significant overhead
• PR #1267: Improve the error message when there is a typo in the column name in the query.
• PR #1261: Remove dimensions from TEXT and FLOAT
• PR #1256: Remove table names from column names for `df
• PR #1253: Collection of fixes for the staging branch
• PR #1246: feat: insertion update index
• PR #1245: Documentation on vector stores + vector benchmark
• PR #1244: feat: create index from projection
• PR #1233: GitHub Data Source Integration
• PR #1115: Add support for Neuralforecast
• PR #1241: Bump Version to v0.3.8+dev
• PR #1239: release 0.3.7

v0.3.7

• PR #1239: release 0.3.7
• PR #1238: Revert "add stable diffusion integration"
• PR #1111: add stable diffusion integration
• PR #1228: fix: improve messaging and fix bugs
• PR #1230: Force fresh .testmondata cache
• PR #1229: Fix staging CI setup
• PR #1225: fix: merging dataframes with None value
• PR #1223: fix: native db bugs
• PR #1209: feat: add testmon in pytest
• PR #1201: Improve Documentation For Model Training
• PR #1210: chore
• PR #1212: fix: 1199 enhance similarity search test to make index is used
• PR #1211: fix: evadb_client fails on launch
• PR #1203: fix: chatgpt exact cache
• PR #1208: fix: if create table fails, we were not cleaning the entry in catalog
• PR #1204: feat: print error msg if creating built-in functions failed
• PR #1180: Adding SET statement for configuration management
• PR #1198: Chroma vector store
• PR #1162: Adding support for Sklearn linear regression in EvaDB
• PR #1196: bug: fix links in model training notebook
• PR #1167: Model training Colab Notebook
• PR #1185: Add support for function calls without input arguments: `FUNC
• PR #1192: Skip pinecone test if pinecone is not installed
• PR #1191: fix: unclosed file error
• PR #1190: fix: release change log
• PR #1153: feat: integrate with pgvector
• PR #1135: Pinecone vector store
• PR #1165: bug: drop bug fix
• PR #1168: LLM app fixes
• PR #1182: Bump Version to v0.3.7+dev
• PR #1181: releass: bump a version further to skip cached wheel

v0.3.6

• PR #1181: releass: bump a version further to skip cached wheel
• PR #1179: fix: release change log
• PR #1177: fix: release action
• PR #1176: fix: release
• PR #1174: fix: trigger build
• PR #1173: fix: release on success only and credential to pass protection rule
• PR #1172: Release v0.3.5
• PR #1169: fix: hotfix for the failing staging build
• PR #1159: Load query exception for invalid file format
• PR #1164: Fix model train with Ludwig on Colab
• PR #1155: feat: add support for enabling ORDER BY on non projected columns
• PR #1158: Adding Algolia search to Eva-DB Docs
• PR #1146: CREATE OR REPLACE FUNCTION
• PR #1154: feat: add github actions to sync and release wheel
• PR #1123: Updates evadb apps
• PR #1157: chore
• PR #977: chore
• PR #1137: feat: add support for if not exists in create db
• PR #1132: docs: add home sale forecast usecase into documentation
• PR #1136: feat: support order by using the projection columns
• PR #1030: Add model inference documentation
• PR #1134: ci: staging build fix
• PR #1124: fix: third-party test
• PR #1118: Add a model forecasting notebook in tutorials
• PR #1125: feat: create table in integration table from evadb select query
• PR #1122: fix: flaky ci unit tests
• PR #1113: fix: update docs and notebooks
• PR #1114: feat: Improve db integration
• PR #1108: Set the right output column type for forecast functions
• PR #1107: Added null handling and tests for case insensitive string matching
• PR #1087: Support SELECT expr; which does not require FROM table
• PR #1090: Making Ludwig and HuggingFace case insensitive
• PR #1027: Adding support for MariaDb as backend for EvaDB
• PR #1101: Fix forecasting integration test
• PR #1094: Fixes date and frequency issues in forecasting
• PR #1096: Rename l_plan and p_plan
• PR #1091: fix: evadb is now consistent with lowercase
• PR #1092: feat: Drop database
• PR #1082: feat: create index if exists
• PR #1088: fix: df merging issues when multiple predicates
• PR #1086: Update parameters documentation for forecast
• PR #1084: Fix column name related issue for Forecast functions
• PR #1073: fix: create index from single document
• PR #1080: pandas.DataFrame.fillna is deprecated
• PR #1060: Bump v0.3.5+dev
• PR #1062: Update UDF to function in model-forecasting.rst

✨ EvaDB v0.3.4 ✨

Core Enhancements, New Integrations, Documentation Revamp, Enhanced Developer Experience, & more

Core Enhancements

• EvaDB now includes support for forecasting.

🛠 New Integrations:

• Introducing support for two new data sources: MySQL and Sqlite.

📖 Documentation Revamp:

• We've made several improvements to our documentation to provide an even better developer experience.

Developer Experience:

• Improved reporting of messages for integrations with data sources.
• Query syntax improvement. 1. UDF in EvaDB has been renamed into FUNCTION for better readability. 2. Remove the quotation mark requirements for parameters in CREATE FUNCTION to improve clarity.
• Added support for SELECT Func(*) FROM Table. This enhancement aims to improve the user experience when using a trained model for inference in EvaDB, particularly in cases involving many columns.

What’s Changed

• PR #1057: fix: staging build fix
• PR #1056: fix: update notebook outputs
• PR #1053: docs: AI powered join
• PR #1054: fix: lint for notebooks
• PR #1051: fix: create function forecast due to incorrect merging
• PR #1047: fix: update udf to function in notebooks and docs
• PR #1046: fix: add missing needed file
• PR #1043: Hot fix forecasting
• PR #1026: Removing quotes from udf_metadata_key
• PR #969: Forecasting in EVA
• PR #1034: feat: UDF migrates to Function
• PR #1035: Fix all doc issues
• PR #1029: Fix the link in new-data-source.rst
• PR #1023: Feat : Sqlite Database Handler
• PR #1019: feat: add CI for MySQL
• PR #1025: Add mysql as an available data source into documentation
• PR #965: msyql integration added
• PR #1018: fix: install pkg and ignore tutorials
• PR #1016: Revamp getting started.
• PR #1014: Extend the circle ci link check to include README.md
• PR #1015: Enable Coverage Check
• PR #1008: fix: CREATE TABLE persists entry in the catalog even if the query fails
• PR #1007: fix: create Table fails when projecting column from native table
• PR #1006: fix: update more verbose message for creating database
• PR #991: feat: add use case of LLM on food review running on colab
• PR #997: fix: improve error msg
• PR #996: Fix all BUILD warning and link validation in the documentation
• PR #986: docs: Update README.md
• PR #985: Bump v0.3.4+dev
• PR #983: doc: fix nav bar

Thanks to @jiashenC, @xzdandy, @rohithmulumudy, @Chitti-Ankith, @americast, @hershd23, @gaurav274, and @jarulraj for contributions!

v0.3.3

• PR #983: doc: fix nav bar
• PR #982: fix: batch merge causing redundant row
• PR #981: fix: use the same interface
• PR #979: docs: added logo
• PR #980: docs: Update README.md
• PR #975: Simplify the ludwig dependency
• PR #972: feat: improve dev doc
• PR #971: Revert "feat: Integrating thirdy party Slack API "
• PR #967: feat: Integrating thirdy party Slack API
• PR #966: Developer guide for new structure data source integration
• PR #949: feat: improve circle ci
• PR #946: Support `SELECT Func
• PR #963: Bump v0.3.3+dev
• PR #962: release 0.3.2

✨ EvaDB v0.3.2 ✨

Core Enhancements, Documentation Revamp, New Integration, & more

Core Enhancements

• 🤖 Introducing direct model training and tuning right inside your database system.
• 📚 Add support to connect with popular native data sources through EvaDB seamlessly. Unleash the power of AI models.

Documentation Revamp 📖

• Several improvements to documentation to enhance your developer experience

Enhanced Usecase 🚀

• Elevate your YouTube QA app blog generator with enhanced capabilities.

New Integrations 🛠

• Postgres Integration: Run native Postgres query through EvaDB, Directly run EvaDB AI query on Postgres native data.
• Ludwig Integration, AutoML on your structure and unstructured data. Train a prediction model through EvaDB from your data sources. Use the trained models as functions to compose customized queries.

Benchmark

• EvaDB runs a text summarization AI task on 12,000 CNN news articles seven times faster than MindsDB.

What’s Changed

• PR #962: release 0.3.2
• PR #953: docs: Fix User Reference and Dev Guide
• PR #952: docs: Update PULL_REQUEST_TEMPLATE.md
• PR #951: docs: updated overview and usecases
• PR #950: docs: improve the docs
• PR #943: Auto drop row id for `CREATE UDF xxx FROM
• PR #945: fix: circle ci pip + setuptools conflict
• PR #942: feat: select from native database
• PR #921: fix: more detailed YouTube QA app blog generator
• PR #940: feat: update ci test for third party integration
• PR #941: chore
• PR #935: Ludwig-based model train and tune support.
• PR #937: feat: integration for pg_handler added
• PR #939: Hot Fix for Fuzzy Join
• PR #936: Use statement
• PR #938: pydantic > 2 conflicts with ray
• PR #932: feat: postgres integration - create database
• PR #931: Revert "Revert "Fix benchmark documentation
• PR #931: Revert "Fix benchmark documentation
• PR #931: Fix benchmark documentation
• PR #930: Text summarization benchmark with MindsDB and EvaDB
• PR #929: Fix problems found when running hugging face text summariztion model on large input.
• PR #928: feat: improve the use case doc
• PR #927: chore
• PR #925: chore
• PR #926: chore
• PR #924: fix tutorials 11-similarity-search-for-motif-mining.ipynb
• PR #922: feat : Run EvaDB on Postgres
• PR #919: chore: redirect to evadb
• PR #918: docs: updates
• PR #916: Bump v0.3.2+dev
• PR #915: Release 0.3.1

Thanks to @jiashenC, @xzdandy, @gaurav274, @jarulraj, @yulaicui, and @Chitti-Ankith for contributions!

v0.3.1

• PR #913: fix: motif_mining query
• PR #895: fix: Renaming Cursor Functions
• PR #911: Added comments for ChatGPT UDF.
• PR #908: feat: migrate to sqlalchemy 2.0
• PR #910: Bump v0.3.1+dev
• PR #907: feat: table filtering based on multiple keywords UDF

v0.3.0

• PR #907: feat: table filtering based on multiple keywords UDF
• PR #906: ci: add 3.11
• PR #905: feat: remove ocr
• PR #904: feat: chatgpt prompt arg
• PR #903: Bump v0.3.0+dev

v0.2.15

• PR #898: fix: index creation better error msg
• PR #817: fix: GPU ids and degree of parallelism
• PR #897: feat: add support for 3.11
• PR #885: feat: pandas qa sample app
• PR #896: app: youtube channel qa app
• PR #893: feat: cleanup create mat view
• PR #892: feat: update notebooks
• PR #894: fix: Youtube app