1.2.10

• 💿 #2430 - Further development of experimental redis support using vectors for text similarity.

Sponsorships

🎉 A huge thanks to LambdaTest for sponsoring ChatterBot! 🎉

1.2.9

• 🐞 #2425 Fix date parsing case
• 🐜 #2426 Add check for sqlite WAL mode status and execute outside of transaction
• 🏋🏼 #2427 Improve bulk operations in training methods
• 📦 #2428 Dependency updates for redis, ollama, and pymongo

1.2.8

• 👣 #2421 Reduce running time of test suite
• 📜 #2422 Django documentation updates and swappable model improvements
• 🥡 #2423 Allow extra args to be passed to the mongodb adapter
• 🐍 #2406 Python 3.13 support!

1.2.7

• 🧪 #2407 Experimental support for large language models (https://docs.chatterbot.us/large-language-models/)
• ⬆️ 8d27a8b Pin upper bounds for the chatterbot-corpus package. Starting in chatterbot-corpus >= 1.3.0 strings longer than 255 characters will be present.

1.2.6

• 296809e Fix for postgres compatibility for the SQLStorageAdapter
• #2412 Generate response search index when in_response_to key is present in files (such as json files passed to the .train() method when using the JsonFileTrainer

1.2.5

• #2408 Django migrations (added help_text attributes to all model fields for the Django admin)
• #2408 Various Django-related bug fixes
• #2410 Fix for loading single files specified by the JsonFileTrainer and CsvFileTrainer

1.2.4

• #2405 Add support for Python versions 3.10, 3.11, 3.12
• #2403 Additional support for loading data from CSV, TSV, and JSON formats.
• #2404 Improved documentation around taggers and more descriptive exception messages
• 387a689 Add a --help flag to python -m chatterbot invocations

1.2.3

• #2397 - Add Redis storage adapter with support for vectors 📈 (docs: https://docs.chatterbot.us/storage/redis/)
• #2398 SpecificResponseAdapter: Accept patterns using spaCy's rule-based matching: https://spacy.io/usage/rule-based-matching
• #2398 SpecificResponseAdapter: Accept a function for the output_text parameter that can be called to generate the output text.

1.2.2

• #2395 - Performance improvements by centralizing search-text generation
• #2395 - Miscellaneous fixes related to searching and matching for response generation

1.2.1

• Use bulk processing from spacy pipeline for improved performance #2392
• Modify TimeLogicAdapter to use SpaCy PhraseMatcher instead of NLTK NaiveBayesClassifier #2394