github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 08:59:13 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #5762] How to pass a custom file (txt\doc\pdf) to prompt through the API #3588

New Issue
Closed
opened 2026-04-12 14:19:37 -05:00 by GiteaMirror · 7 comments
GiteaMirror commented 2026-04-12 14:19:37 -05:00
Owner
Copy Link

Originally created by @liukesoft on GitHub (Jul 18, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/5762

Hello, I have deployed Olama on my local Linux and would like to call it through an API. The model is llama3, and prompt is my custom Python template, which allows the model to generate new Python code according to my Python template. How can I send my template (txt \ doc \ pdf) to the interface through API calls? How to set parameters? thanks！！

curl http://localhost:11434/api/generate -d '{
"model": "llama3",
"prompt":"Why is the sky blue?" //How to use custom prompt templates？？
}'

Originally created by @liukesoft on GitHub (Jul 18, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/5762 Hello, I have deployed Olama on my local Linux and would like to call it through an API. The model is llama3, and prompt is my custom Python template, which allows the model to generate new Python code according to my Python template. How can I send my template (txt \ doc \ pdf) to the interface through API calls? How to set parameters? thanks！！ curl http://localhost:11434/api/generate -d '{ "model": "llama3", "prompt":"Why is the sky blue?" //How to use custom prompt templates？？ }'
GiteaMirror added the question label 2026-04-12 14:19:37 -05:00
GiteaMirror commented 2026-04-12 14:19:38 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jul 18, 2024):

ollama is just an inference engine, it doesn't do document extraction. For that you would use something like a document loader from langchain_community.document_loaders or llama_parse.LLamaParse. Then you include the extracted information along with your prompt in the prompt field of the message you send to ollama. The parameters you can set are laid out in the API guide.

<!-- gh-comment-id:2237027537 --> @rick-github commented on GitHub (Jul 18, 2024): ollama is just an inference engine, it doesn't do document extraction. For that you would use something like a document loader from `langchain_community.document_loaders` or `llama_parse.LLamaParse`. Then you include the extracted information along with your prompt in the `prompt` field of the message you send to ollama. The parameters you can set are laid out in the [API guide](https://github.com/ollama/ollama/blob/main/docs/api.md).
GiteaMirror commented 2026-04-12 14:19:39 -05:00
Author
Owner
Copy Link

@robotom commented on GitHub (Oct 24, 2024):

I can do the text extraction from the document. At that point, is it recommended to just pass entire walls of text into each prompt?

Are there any tunable settings to maximize the model's understanding of this wall of text?

Or is it just as simple of dumping all the text into the message that I send to the model.

Any advice on this would be super helpful! Thank you.

<!-- gh-comment-id:2436489660 --> @robotom commented on GitHub (Oct 24, 2024): I can do the text extraction from the document. At that point, is it recommended to just pass entire walls of text into each prompt? Are there any tunable settings to maximize the model's understanding of this wall of text? Or is it just as simple of dumping all the text into the message that I send to the model. Any advice on this would be super helpful! Thank you.
GiteaMirror commented 2026-04-12 14:19:39 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 24, 2024):

Increase the context size (https://github.com/ollama/ollama/issues/7341) and send the text.

<!-- gh-comment-id:2436492711 --> @rick-github commented on GitHub (Oct 24, 2024): Increase the context size (https://github.com/ollama/ollama/issues/7341) and send the text.
GiteaMirror commented 2026-04-12 14:19:39 -05:00
Author
Owner
Copy Link

@robotom commented on GitHub (Oct 24, 2024):

Thanks. Any idea on how well the model will understand this text? Are there other methods like tuning, params, etc. to achieve this?

Is the max recommended 4096? Are there any larger numbers, etc. Is there any data on this relationship to CPU/RAM, etc?

Thanks.

<!-- gh-comment-id:2436503583 --> @robotom commented on GitHub (Oct 24, 2024): Thanks. Any idea on how well the model will understand this text? Are there other methods like tuning, params, etc. to achieve this? Is the max recommended 4096? Are there any larger numbers, etc. Is there any data on this relationship to CPU/RAM, etc? Thanks.
GiteaMirror commented 2026-04-12 14:19:40 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 24, 2024):

https://github.com/ollama/ollama/issues/7341#issuecomment-2436504435

<!-- gh-comment-id:2436507330 --> @rick-github commented on GitHub (Oct 24, 2024): https://github.com/ollama/ollama/issues/7341#issuecomment-2436504435
GiteaMirror commented 2026-04-12 14:19:40 -05:00
Author
Owner
Copy Link

@armartinez commented on GitHub (Sep 24, 2025):

@rick-github The ollama app supports passing files as input, is this functionality available on the API ?

<!-- gh-comment-id:3328147879 --> @armartinez commented on GitHub (Sep 24, 2025): @rick-github The ollama app supports passing files as input, is this functionality available on the API ?
GiteaMirror commented 2026-04-12 14:19:41 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 24, 2025):

The app does as outlined in https://github.com/ollama/ollama/issues/5762#issuecomment-2237027537: it converts the input file into text that is added to the prompt. The API doesn't provide this, it's done by the client. In the app's case, the app is the client doing the conversion.

<!-- gh-comment-id:3328164659 --> @rick-github commented on GitHub (Sep 24, 2025): The app does as outlined in https://github.com/ollama/ollama/issues/5762#issuecomment-2237027537: it converts the input file into text that is added to the prompt. The API doesn't provide this, it's done by the client. In the app's case, the app is the client doing the conversion.
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 question
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#3588
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?