github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-21 13:11:39 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #15933] Model Proposal: CAJAL - Scientific Paper Generation Modelfile #72207

New Issue
Open
opened 2026-05-05 03:38:02 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-05-05 03:38:02 -05:00
Owner
Copy Link

Originally created by @Agnuxo1 on GitHub (May 2, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/15933

Proposal: Add CAJAL to Ollama Model Library

What: A specialized modelfile for generating publication-ready scientific papers.

Why this fits Ollama:

  • Ollama enables local AI — researchers using Ollama need academic writing capabilities
  • CAJAL generates 7-section papers with real arXiv citations via structured prompts
  • Zero additional dependencies — pure prompt engineering + any base model
  • MIT licensed, open source, community-driven

What CAJAL provides:

  • System prompt optimized for scientific writing
  • Structured generation (Abstract → Intro → Methods → Results → Discussion → Conclusion → References)
  • Tribunal scoring (8-10 judges evaluate quality)
  • Works with llama3.1, qwen3.5, or any Ollama model

Implementation: https://github.com/Agnuxo1/CAJAL/tree/main/integrations/ollama/Modelfile

Main repo: https://github.com/Agnuxo1/CAJAL

Happy to adapt to Ollama's modelfile conventions and library standards.

Originally created by @Agnuxo1 on GitHub (May 2, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/15933 ## Proposal: Add CAJAL to Ollama Model Library **What:** A specialized modelfile for generating publication-ready scientific papers. **Why this fits Ollama:** - Ollama enables local AI — researchers using Ollama need academic writing capabilities - CAJAL generates 7-section papers with real arXiv citations via structured prompts - Zero additional dependencies — pure prompt engineering + any base model - MIT licensed, open source, community-driven **What CAJAL provides:** - System prompt optimized for scientific writing - Structured generation (Abstract → Intro → Methods → Results → Discussion → Conclusion → References) - Tribunal scoring (8-10 judges evaluate quality) - Works with llama3.1, qwen3.5, or any Ollama model **Implementation:** https://github.com/Agnuxo1/CAJAL/tree/main/integrations/ollama/Modelfile **Main repo:** https://github.com/Agnuxo1/CAJAL Happy to adapt to Ollama's modelfile conventions and library standards.
GiteaMirror commented 2026-05-05 03:38:03 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 3, 2026):

The implementation link is 404.

Ollama doesn't generally host Modelfiles, if you want to provide a link for a wider audience, create a PR to add it to the integrations.

If you want to provide a model that provides a CAJAL service, create it and push it to the user library.

<!-- gh-comment-id:4367108502 --> @rick-github commented on GitHub (May 3, 2026): The implementation link is 404. Ollama doesn't generally host Modelfiles, if you want to provide a link for a wider audience, create a PR to add it to the [integrations](https://github.com/ollama/ollama#community-integrations). If you want to provide a model that provides a CAJAL service, create it and [push](https://github.com/ollama/ollama/blob/main/docs/import.mdx#sharing-your-model-on-ollamacom) it to the user library.
GiteaMirror commented 2026-05-05 03:38:03 -05:00
Author
Owner
Copy Link

@Agnuxo1 commented on GitHub (May 4, 2026):

@rick-github Thanks for the clarification! You're right — the Modelfile link needs fixing (I see the path changed after reorganization).

I'll take both paths you suggested:

  1. Fix the Modelfile link in our repo and prepare a PR to add CAJAL to the Ollama integrations page (or community integrations repo if that's the right target).

  2. Publish CAJAL to the Ollama user library — I'll create a public model entry so users can run ollama pull cajal directly.

Quick question: should the integration PR go to ollama/ollama under docs/ or is there a separate ollama/integrations repository? Want to make sure I target the right place.

Appreciate the guidance — this kind of feedback is exactly what helps us improve the proposal.

— Francisco (@Agnuxo1), P2PCLAW

<!-- gh-comment-id:4372442397 --> @Agnuxo1 commented on GitHub (May 4, 2026): @rick-github Thanks for the clarification! You're right — the Modelfile link needs fixing (I see the path changed after reorganization). I'll take both paths you suggested: 1. **Fix the Modelfile link** in our repo and prepare a **PR to add CAJAL to the Ollama integrations page** (or community integrations repo if that's the right target). 2. **Publish CAJAL to the Ollama user library** — I'll create a public model entry so users can run `ollama pull cajal` directly. Quick question: should the integration PR go to `ollama/ollama` under `docs/` or is there a separate `ollama/integrations` repository? Want to make sure I target the right place. Appreciate the guidance — this kind of feedback is exactly what helps us improve the proposal. — Francisco (@Agnuxo1), P2PCLAW
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
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#72207
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?