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 #857] Manually download and upload models #62450

New Issue
Closed
opened 2026-05-03 09:00:54 -05:00 by GiteaMirror · 12 comments
GiteaMirror commented 2026-05-03 09:00:54 -05:00
Owner
Copy Link

Originally created by @dawnpatrol04 on GitHub (Oct 20, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/857

hey guys. Having issues getting with this part a work with corporate proxy: docker exec -it ollama ollama run llama2.

2 issues.

  1. When I set a proxy something breaks.

  2. model url / cert not allowed / blocked. To work around this I will need to manually download model files upload to the container.

Can we manually download and upload model files? Where do I put the model files after I have download them?

Originally created by @dawnpatrol04 on GitHub (Oct 20, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/857 hey guys. Having issues getting with this part a work with corporate proxy: docker exec -it ollama ollama run llama2. 2 issues. 1) When I set a proxy something breaks. 2) model url / cert not allowed / blocked. To work around this I will need to manually download model files upload to the container. Can we manually download and upload model files? Where do I put the model files after I have download them?
GiteaMirror added the bug label 2026-05-03 09:00:54 -05:00
GiteaMirror commented 2026-05-03 09:01:11 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Oct 25, 2023):

You can set HTTP_PROXY or HTTPS_PROXY when starting the ollama docker container.

docker run -d -e HTTP_PROXY -e HTTPS_PROXY ollama/ollama

You can also set it globally for all docker contains. See https://docs.docker.com/network/proxy/

<!-- gh-comment-id:1779871058 --> @mxyng commented on GitHub (Oct 25, 2023): You can set `HTTP_PROXY` or `HTTPS_PROXY` when starting the ollama docker container. ``` docker run -d -e HTTP_PROXY -e HTTPS_PROXY ollama/ollama ``` You can also set it globally for all docker contains. See https://docs.docker.com/network/proxy/
GiteaMirror commented 2026-05-03 09:01:13 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Oct 30, 2023):

Hi @dawnpatrol04, as @mxyng mentioned this may be from not configuring Ollama to use your http or https proxy. If this doesn't fix the issue please feel free to re-open!

<!-- gh-comment-id:1786121973 --> @jmorganca commented on GitHub (Oct 30, 2023): Hi @dawnpatrol04, as @mxyng mentioned this may be from not configuring Ollama to use your http or https proxy. If this doesn't fix the issue please feel free to re-open!
GiteaMirror commented 2026-05-03 09:01:17 -05:00
Author
Owner
Copy Link

@reactivetype commented on GitHub (Nov 3, 2023):

@dawnpatrol04 @jmorganca this issue is still not solved. There many open issues about proxy and it'd be great if you provide a way and documentation to download the models manually.

<!-- gh-comment-id:1792864180 --> @reactivetype commented on GitHub (Nov 3, 2023): @dawnpatrol04 @jmorganca this issue is still not solved. There many open issues about proxy and it'd be great if you provide a way and documentation to download the models manually.
GiteaMirror commented 2026-05-03 09:01:18 -05:00
Author
Owner
Copy Link

@nmstoker commented on GitHub (Feb 18, 2024):

I second reactivetype's comment. Can this be reopened?

It would be wise to park the proxy point completely and just confirm if there's a way to download the models manually as this is often necessary in corporate networks, some of which will have areas where there is literally no internet access allowed.

<!-- gh-comment-id:1951406977 --> @nmstoker commented on GitHub (Feb 18, 2024): I second reactivetype's comment. Can this be reopened? It would be wise to park the proxy point completely and just **confirm if there's a way to download the models manually** as this is often necessary in corporate networks, some of which will have areas where there is literally no internet access allowed.
GiteaMirror commented 2026-05-03 09:01:20 -05:00
Author
Owner
Copy Link

@nmstoker commented on GitHub (Feb 18, 2024):

I'm guessing the answer may involve putting files in the location here:

https://github.com/ollama/ollama/blob/main/docs/faq.md#where-are-models-stored

I'll see if I can figure it out from common sense and a bit of tinkering

<!-- gh-comment-id:1951409054 --> @nmstoker commented on GitHub (Feb 18, 2024): I'm guessing the answer may involve putting files in the location here: https://github.com/ollama/ollama/blob/main/docs/faq.md#where-are-models-stored I'll see if I can figure it out from common sense and a bit of tinkering
GiteaMirror commented 2026-05-03 09:01:21 -05:00
Author
Owner
Copy Link

@dhirajsuvarna commented on GitHub (Mar 4, 2024):

did someone figure this out?

<!-- gh-comment-id:1975787178 --> @dhirajsuvarna commented on GitHub (Mar 4, 2024): did someone figure this out?
GiteaMirror commented 2026-05-03 09:01:22 -05:00
Author
Owner
Copy Link

@sven700c commented on GitHub (Apr 5, 2024):

I second reactivetype's comment. Can this be reopened?

It would be wise to park the proxy point completely and just confirm if there's a way to download the models manually as this is often necessary in corporate networks, some of which will have areas where there is literally no internet access allowed.

Exactly this! There is not Internet access allowed at all on the machine I'm trying to run this and other machines only get https access thru a web browser, so setting https_proxy env var for the command line won't help.

<!-- gh-comment-id:2038767065 --> @sven700c commented on GitHub (Apr 5, 2024): > I second reactivetype's comment. Can this be reopened? > > It would be wise to park the proxy point completely and just **confirm if there's a way to download the models manually** as this is often necessary in corporate networks, some of which will have areas where there is literally no internet access allowed. Exactly this! There is not Internet access allowed at all on the machine I'm trying to run this and other machines only get https access thru a web browser, so setting https_proxy env var for the command line won't help.
GiteaMirror commented 2026-05-03 09:01:23 -05:00
Author
Owner
Copy Link

@nmstoker commented on GitHub (Apr 5, 2024):

@jmorganca any chance this could be re-opened & looked at again purely in regard to manual downloading of models for the reasons given above?

<!-- gh-comment-id:2039239659 --> @nmstoker commented on GitHub (Apr 5, 2024): @jmorganca any chance this could be re-opened & looked at again purely in regard to manual downloading of models for the reasons given above?
GiteaMirror commented 2026-05-03 09:01:24 -05:00
Author
Owner
Copy Link

@Seedmanc commented on GitHub (Apr 12, 2024):

I'm not even in a corporative network and pulling model doesn't work at all. I wish someone made a torrent of a working Ollama installation so I would at least have something to start with, driven by example. Starting from scratch is too hard, not a single model comes with the installer itself.

<!-- gh-comment-id:2051898176 --> @Seedmanc commented on GitHub (Apr 12, 2024): I'm not even in a corporative network and pulling model doesn't work at all. I wish someone made a torrent of a working Ollama installation so I would at least have something to start with, driven by example. Starting from scratch is too hard, not a single model comes with the installer itself.
GiteaMirror commented 2026-05-03 09:01:26 -05:00
Author
Owner
Copy Link

@MattFriedman commented on GitHub (May 5, 2024):

I too and behind a corporate proxy. I've set the http mentioned above env vars and have no luck getting ollama to use my proxy. On the other hand, I can use curl with my proxy to communicate the the external endpoints so for certain the proxy works perfectly. I've tried running in docker, running from the cli and running from the source code. Constantly getting a bad gateway error.

I would like to download the model manually and put it in the right location if someone knows how to do that.

<!-- gh-comment-id:2094848148 --> @MattFriedman commented on GitHub (May 5, 2024): I too and behind a corporate proxy. I've set the http mentioned above env vars and have no luck getting ollama to use my proxy. On the other hand, I can use curl with my proxy to communicate the the external endpoints so for certain the proxy works perfectly. I've tried running in docker, running from the cli and running from the source code. Constantly getting a bad gateway error. I would like to download the model manually and put it in the right location if someone knows how to do that.
GiteaMirror commented 2026-05-03 09:01:28 -05:00
Author
Owner
Copy Link

@usathyan commented on GitHub (Jun 11, 2024):

it downloads the files as manifests and blobs. not easy to do manually. this is a great useless product unfortunately for corporate uses.

<!-- gh-comment-id:2160562079 --> @usathyan commented on GitHub (Jun 11, 2024): it downloads the files as manifests and blobs. not easy to do manually. this is a great useless product unfortunately for corporate uses.
GiteaMirror commented 2026-05-03 09:01:29 -05:00
Author
Owner
Copy Link

@amirrezaDev1378 commented on GitHub (Aug 28, 2024):

If anyone did not find a solution for their problem, I've created this simple app that will give you links to download and model in any size you want from the Ollama registry:

https://github.com/amirrezaDev1378/ollama-model-direct-download

<!-- gh-comment-id:2316085490 --> @amirrezaDev1378 commented on GitHub (Aug 28, 2024): If anyone did not find a solution for their problem, I've created this simple app that will give you links to download and model in any size you want from the Ollama registry: https://github.com/amirrezaDev1378/ollama-model-direct-download
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 bug
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#62450
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?