github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 05:05:37 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #3510] Databases #79841

New Issue
Open
opened 2026-05-09 07:01:46 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-05-09 07:01:46 -05:00
Owner
Copy Link

Originally created by @trymeouteh on GitHub (Apr 6, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/3510

What are you trying to do?

A way to manage, and create and train databases that can run on-top of models

How should we solve this?

Docker has images, containers and volumes. Ollama currently only has models.

To add a new category for databases. Databases can be managed by...

  • Import from file
  • Export to file
  • Download from registry
  • Uploaded to registry
  • Create new empty database
  • Rename database
  • Delete database
  • Duplicate database (Copy/paste database)

Databases can be edited by training the databases on files (documents, images, videos, audios, web pages, etc). A Ollama GUI would be best for training a database.

Now you have models and databases, when you run a model, you can add one or more databases with it which it will use as additional trained data.

What is the impact of not solving this?

Ollama will be limited to LLMs and unable to add additional training data on-top of LLMs.

Anything else?

Adding an additional feature when running models to choose a database to send any trained data from prompt. If a response is good, user can send a like (thumbs up) to the database to train the model that the response was good. If a response is bad, user can send a dislike (thumbs down) to the database to train the model that the response was bad.

Adding a tagging feature to databases would be useful
https://github.com/ollama/ollama/issues/2847

Originally created by @trymeouteh on GitHub (Apr 6, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/3510 ### What are you trying to do? A way to manage, and create and train databases that can run on-top of models ### How should we solve this? Docker has images, containers and volumes. Ollama currently only has models. To add a new category for databases. Databases can be managed by... - Import from file - Export to file - Download from registry - Uploaded to registry - Create new empty database - Rename database - Delete database - Duplicate database (Copy/paste database) Databases can be edited by training the databases on files (documents, images, videos, audios, web pages, etc). A Ollama GUI would be best for training a database. Now you have models and databases, when you run a model, you can add one or more databases with it which it will use as additional trained data. ### What is the impact of not solving this? Ollama will be limited to LLMs and unable to add additional training data on-top of LLMs. ### Anything else? Adding an additional feature when running models to choose a database to send any trained data from prompt. If a response is good, user can send a like (thumbs up) to the database to train the model that the response was good. If a response is bad, user can send a dislike (thumbs down) to the database to train the model that the response was bad. Adding a tagging feature to databases would be useful https://github.com/ollama/ollama/issues/2847
GiteaMirror added the feature request label 2026-05-09 07:01:46 -05:00
GiteaMirror commented 2026-05-09 07:01:48 -05:00
Author
Owner
Copy Link

@AncientMystic commented on GitHub (May 9, 2024):

I would really love to see this as well as some way to bulk add documents via a directory / set watch folders to databases to be slowly processed / indexed in the background as i have many documents i would love to combine with LLMs, this would be the greatest feature ollama could possibly add.

it would also open up the possibility of learning more from user input/chats and recall for conversations to build on data and give more long term user specific context and personalisation to responses

<!-- gh-comment-id:2101815148 --> @AncientMystic commented on GitHub (May 9, 2024): I would really love to see this as well as some way to bulk add documents via a directory / set watch folders to databases to be slowly processed / indexed in the background as i have many documents i would love to combine with LLMs, this would be the greatest feature ollama could possibly add. it would also open up the possibility of learning more from user input/chats and recall for conversations to build on data and give more long term user specific context and personalisation to responses
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#79841
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?