github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 16:40:08 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #7755] Proper way to train model on my data and load into Ollama? #67008

New Issue
Closed
opened 2026-05-04 09:14:00 -05:00 by GiteaMirror · 4 comments
GiteaMirror commented 2026-05-04 09:14:00 -05:00
Owner
Copy Link

Originally created by @robotom on GitHub (Nov 20, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7755

As I've mentioned in the title, I have some huge text based documents which exceed typical context windows, even on large machines with large models (e.g. 405B). Is there a way I could train llama 3.1:8B (for example) on these docs and then load it into Ollama and ask the model about them? Thank you!

Originally created by @robotom on GitHub (Nov 20, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7755 As I've mentioned in the title, I have some huge text based documents which exceed typical context windows, even on large machines with large models (e.g. 405B). Is there a way I could train llama 3.1:8B (for example) on these docs and then load it into Ollama and ask the model about them? Thank you!
GiteaMirror added the feature request label 2026-05-04 09:14:00 -05:00
GiteaMirror commented 2026-05-04 09:14:03 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Nov 20, 2024):

Your options are RAG or fine-tuning. There are RAG implementations for ollama in the community integrations. RAG doesn't "understand" the document, though, it just searches based on a query and responds with a result synthesized from the search results. Fine-tuning options include unsloth, llama factory and axolotl. The problem with fine-tuning is it takes time to generate the new model, if you are receiving daily doc dumps it might not be fit for purpose. A new entry in to the scene is instructlab, which from the sound of it is a more iterative approach to fine-tuning, so might fit your use case better. I haven't used it yet so my understanding may be incorrect.

<!-- gh-comment-id:2487704178 --> @rick-github commented on GitHub (Nov 20, 2024): Your options are [RAG](https://en.wikipedia.org/wiki/Retrieval-augmented_generation) or [fine-tuning](https://en.wikipedia.org/wiki/Fine-tuning_(deep_learning)). There are RAG implementations for ollama in the [community integrations](https://github.com/ollama/ollama?tab=readme-ov-file#community-integrations). RAG doesn't "understand" the document, though, it just searches based on a query and responds with a result synthesized from the search results. Fine-tuning options include [unsloth](https://unsloth.ai/), [llama factory](https://github.com/hiyouga/LLaMA-Factory) and [axolotl](https://github.com/axolotl-ai-cloud/axolotl). The problem with fine-tuning is it takes time to generate the new model, if you are receiving daily doc dumps it might not be fit for purpose. A new entry in to the scene is [instructlab](https://github.com/instructlab), which from the sound of it is a more iterative approach to fine-tuning, so might fit your use case better. I haven't used it yet so my understanding may be incorrect.
GiteaMirror commented 2026-05-04 09:14:03 -05:00
Author
Owner
Copy Link

@robotom commented on GitHub (Nov 20, 2024):

Thanks for that. I had a look at the links for fine-tuning. Llama factory looks simple enough to use. I've never done this before but I think it's something like this:

  1. Create a "dataset" --- not sure what format this needs to be in but I have a bunch of docs/txts.
  2. Upload dataset, tweak the parameters and press "Train/Start".
  3. Load the new_llama3_model into an engine --- I want to load it into Ollama locally on my machine.

Do you have any advice on 1-2-3? I am designing a piece of software and I really want it to revolve around Ollama as the underlying engine.

I also don't mind taking some time to generate new models. I take it we're talking days not weeks of GPU time.

Thanks!

<!-- gh-comment-id:2487891201 --> @robotom commented on GitHub (Nov 20, 2024): Thanks for that. I had a look at the links for fine-tuning. Llama factory looks simple enough to use. I've never done this before but I think it's something like this: 1) Create a "dataset" --- not sure what format this needs to be in but I have a bunch of docs/txts. 2) Upload dataset, tweak the parameters and press "Train/Start". 3) Load the new_llama3_model into an engine --- I want to load it into Ollama locally on my machine. Do you have any advice on 1-2-3? I am designing a piece of software and I really want it to revolve around Ollama as the underlying engine. I also don't mind taking some time to generate new models. I take it we're talking days not weeks of GPU time. Thanks!
GiteaMirror commented 2026-05-04 09:14:03 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 13, 2025):

https://www.datacamp.com/tutorial/llama-factory-web-ui-guide-fine-tuning-llms

<!-- gh-comment-id:2586024704 --> @rick-github commented on GitHub (Jan 13, 2025): https://www.datacamp.com/tutorial/llama-factory-web-ui-guide-fine-tuning-llms
GiteaMirror commented 2026-05-04 09:14:04 -05:00
Author
Owner
Copy Link

@OussamaAdaoumoum commented on GitHub (Apr 7, 2025):

@robotom , could I have your email or another way to contact you to ask about this, please?

<!-- gh-comment-id:2782920030 --> @OussamaAdaoumoum commented on GitHub (Apr 7, 2025): @robotom , could I have your email or another way to contact you to ask about this, please?
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#67008
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?