Releases: BBC-Esq/VectorDB-Plugin
Releases · BBC-Esq/VectorDB-Plugin
v7.9.0 - bug fixes
Version 7.9.0 changes (from 7.8.2):
- Fixed vision models (but florence2 still not working)
- Fixed bug whereby HF token was required even for non-gated repositories
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prior version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- Last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specifically:
python -m venv .
.\Scripts\activate
python setup_windows.py
v7.8.2 - BLITZKRIEG
Version 7.8.2 changes (from 7.8.0):
- Updated dependencies, including numpy now version 2+.
- Fixes model info for infly embedding models.
- Fixes batch size for embedding models.
- Fixed a major bug that allowed non-string extracted "text" to slipt through, causing tokenizer errors.
- Fixed file hash being computed twice unnecessarily.
- Fixed the .txt file that TTS relies on from containing thinking.
- Made OCR check optional due to massive processing time with a lot of pdf files.
- Improved OCR check functionality.
- Fixed a bug that required HF authentication for non-gated repos, and threw an error if one wasn't provided.
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prior version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- Last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specifically:
python -m venv .
.\Scripts\activate
python setup_windows.py
v7.8.1 - blitzkrieg
Version 7.8.1 changes (from 7.8.0):
- Updated dependencies, including numpy now version 2+.
- Fixes model info for infly embedding models.
- Fixes batch size for embedding models.
- Fixed a major bug that allowed non-string extracted "text" to slipt through, causing tokenizer errors.
- Fixed file hash being computed twice unnecessarily.
- Fixed the .txt file that TTS relies on from containing thinking.
- Made OCR check optional due to massive processing time with a lot of pdf files.
- Improved OCR check functionality.
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prior version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release v7.8.1 and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- Last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specifically:
python -m venv .
.\Scripts\activate
python setup_windows.py
v7.8.0 - NUMBER ONE
Version 7.8.0 changes (from 7.7.0):
infly-retriever-v1-1.5bis currently the number one MTEB ranked embedding model
Chat Models
Added GLM4Z - 9b - excellent thinking model
Added Phi 4 Mini - 4b
Added Seed Coder - 8b - excellent coding-focused model
Embedding Models
Added gte-Qwen2-1.5B-instruct
Added infly-retriever-v1-1.5b
Added infly-retriever-v1-7b
Removed Alibaba-gte-base - superseded
Removed Alibaba-gte-modernbert-base - superseded
Removed Alibaba-gte-large - superseded
Removed stella_en_1.5B_v5 - superseded
Removed stella_en_400M_v5 - superseded
General Improvements
- Improved the speed and accuracy of adding files that will be put into the database (symlink creation).
- Fine-tuned embedding model key word arguments
- Improve text extraction.
- Correct PDF page citations
- Refactor
gui_tabs_database_query.py - Improve thinking indicators and displays
- Improve scraper
- Refactor
metrics_bar.py - Refactor
module_chat.py
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prior version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release v7.8.0 and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specifically:
python -m venv .
.\Scripts\activate
python setup_windows.py
v7.7.4 - BUG fix
- Bug fix. to Kobold backend. Thanks to @berreka
Contributors
berreka
Assets 2
v7.7.3 - BUG fix
- Fix a bug with LM Studio backend.
- Thanks @berreka
Contributors
berreka
Assets 2
v7.7.2 - QWEN3 DESTROYS
Version 7.7.2 changes:
Vision Models
- Fixed bugs with the
Ovis2andGranite Visionmodels.
Chat Models
Gemma3 models replaced multiple other chat models. They're just too good. They decimated other models.
Embedding Models
- Fix embedding model args
Optical Character Recognition
- Fixed some bugs and improved speed a little bit.
Ask Jeeves
- The Ask Jeeves functionality is outdated at the moment due to data loss issues...will work to repair in a future patch.
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prior version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release v7.7.1 and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specificallyk:
python -m venv .
.\Scripts\activate
python setup_windows.py
v7.7.1 - Qwen3 DESTROYS (patch)
Version 7.7.1 changes:
Vision Models
- Fixed bugs with the
Ovis2andGranite Visionmodels.
Chat Models
Gemma3 models replaced multiple other chat models. They're just too good. They decimated other models.
Embedding Models
Optical Character Recognition
- Fixed some bugs and improved speed a little bit.
Ask Jeeves
- The Ask Jeeves functionality is outdated at the moment due to data loss issues...will work to repair in a future patch.
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prior version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release v7.7.1 and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specificallyk:
python -m venv .
.\Scripts\activate
python setup_windows.py
v7.7.0 - Qwen3 DESTROYS
Version 7.7.0 changes:
Vision Models
- Fixed bugs with the
Ovis2andGranite Visionmodels.
Chat Models
Gemma3 models replaced multiple other chat models. They're just too good. They decimated other models.
Embedding Models
Optical Character Recognition
- Fixed some bugs and improved speed a little bit.
Ask Jeeves
- The Ask Jeeves functionality is outdated at the moment due to data loss issues...will work to repair in a future patch.
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prio 9A64 r version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release v7.7.0 and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specificallyk:
python -m venv .
.\Scripts\activate
python setup_windows.py
v7.6.0 - consolidate stuff
Version 7.5.0 to 7.6.0 changes:
Vision Models
- Fixed bugs with the
Ovis2andGranite Visionmodels.
| Change | Details | Notes |
|---|---|---|
| Removed | Llava Vicuna - 13b |
Superseded by other models with higher quality, speed or both. |
Chat Models
| Change | Details | Notes |
|---|---|---|
| Added | Deepseek R1 - 1.5b |
Generalist thinking model |
| Added | Olympic Coder - 7b |
Experimental coding thinking model focusing on C++ and Python |
| Added | Exaone Deep - 7.8b |
Generalist thinking model |
| Added | OLMo 2 - 13b |
Generalist model |
| Added | Reka Flash - 21b |
Generalist thinking model |
| Added | Exaone Deep - 32b |
Generalist thinking model |
| Added | Olympic Coder - 32b |
Experimental coding thinking model focusing on C++ and Python |
| Added | QwQ - 32b |
Generalist thinking model that also ranks high in coding |
| Removed | Qwen - 32b |
Superseded by Exaone - 32b |
| Removed | Qwen - 7b |
Superseded by Exaone - 7.8b |
| Removed | Qwen - 3b |
Superseded by Exaone - 2.4b |
| Removed | Zephyr 1.6b and Zephyr 3b |
Superseded by newer models |
Embedding Models
| Change | Details | Notes |
|---|---|---|
| Removed | Sentence-t5 models |
Hardly used. |
| Removed | static-retrieval-mrl-en-v1 |
Too difficult to run reliably. |
Optical Character Recognition
- Bug fix to install the correct version of
tesseractbased on whether a user is using Python 3.11 or Python 3.12.
General Improvements
- Updated documentation scraping libraries.
- Updated all dependencies.
- Updated
Jeevesknowledge-base
Installation
- Just follow the installation instructions on the github readme.
Upgrading from a Prior Version
INSTRUCTIONS FOR UPGRADING FROM A PRIOR VERSION
To upgrade from a prior version while still keeping your databases and models that were downloaded, perform the following steps:
- In your directory that contains all of the files currently...delete the
Include,Lib,Scriptsfolders. (pertain to your virtual environment) - Delete the
pyvenv.cfgfile. (pertains to your virtual environment) - At this point, any and all files pertaining to your virtual environment are gone.
- Delete the
AssetsandCSSfolders. (new ones will be provided in their entirety further below) - Extract all the files from the
.zipfile for this release v7.6.0 and go into thesrcfolder. - Copy all files and folders EXCEPT THE CONFIG.YAML FILE into the aforementioned folder where you virtual environment USED TO BE. If there are any prompts indicating that files will be overwritten, CLICK YES since we're only updating necessary files now.
- EXCEPT THE CONFIG.YAML FILE!
- EXCEPT THE CONFIG.YAML FILE!
- last time...EXCEPT THE CONFIG.YAML FILE!
Afterwards, open a command prompt and run the installation instructions on this repository's page; specificallyk:
python -m venv .
.\Scripts\activate
python setup_windows.py