github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #6449] Microsoft Phi-3.5 models #66094

New Issue
Open
opened 2026-05-03 23:57:09 -05:00 by GiteaMirror · 11 comments
GiteaMirror commented 2026-05-03 23:57:09 -05:00
Owner
Copy Link

Originally created by @animaldomestico on GitHub (Aug 20, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6449

Originally created by @animaldomestico on GitHub (Aug 20, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6449 - [Phi-3.5-mini-instruct](https://huggingface.co/microsoft/Phi-3.5-mini-instruct) - [Phi-3.5-MoE-instruct](https://huggingface.co/microsoft/Phi-3.5-MoE-instruct) - [Phi-3.5-vision-instruct](https://huggingface.co/microsoft/Phi-3.5-vision-instruct)
GiteaMirror added the model label 2026-05-03 23:57:10 -05:00
GiteaMirror commented 2026-05-03 23:57:15 -05:00
Author
Owner
Copy Link

@kiradzS commented on GitHub (Aug 21, 2024):

I tried, the mini seems to work fine originally, didn't try the others

<!-- gh-comment-id:2300178310 --> @kiradzS commented on GitHub (Aug 21, 2024): I tried, the mini seems to work fine originally, didn't try the others
GiteaMirror commented 2026-05-03 23:57:17 -05:00
Author
Owner
Copy Link

@coder543 commented on GitHub (Aug 21, 2024):

Yes, that one should “just work”. It’s the other two that will likely be more of a problem… especially the vision one, since I don’t think Phi-3 vision ever got supported, and now there’s a new version of it.

<!-- gh-comment-id:2300281608 --> @coder543 commented on GitHub (Aug 21, 2024): Yes, that one should “just work”. It’s the other two that will likely be more of a problem… especially the vision one, since I don’t think Phi-3 vision *ever* got supported, and now there’s a new version of it.
GiteaMirror commented 2026-05-03 23:57:19 -05:00
Author
Owner
Copy Link

@MathiasSchindler commented on GitHub (Aug 21, 2024):

Using "ollama run phi3.5" works for a short while and then the output turns completely erratic.

<!-- gh-comment-id:2302460907 --> @MathiasSchindler commented on GitHub (Aug 21, 2024): Using "ollama run phi3.5" works for a short while and then the output turns completely erratic.
GiteaMirror commented 2026-05-03 23:57:20 -05:00
Author
Owner
Copy Link

@MeinDeutschkurs commented on GitHub (Aug 22, 2024):

When I use "ollama run phi3.5", the model replies but it does not stop. It just continues with anything else until the token window is full.

<!-- gh-comment-id:2304414234 --> @MeinDeutschkurs commented on GitHub (Aug 22, 2024): When I use "ollama run phi3.5", the model replies but it does not stop. It just continues with anything else until the token window is full.
GiteaMirror commented 2026-05-03 23:57:22 -05:00
Author
Owner
Copy Link

@NightingaleCen commented on GitHub (Aug 22, 2024):

When I use "ollama run phi3.5", the model replies but it does not stop. It just continues with anything else until the token window is full.

Might be related to ggerganov/llama.cpp#9127 ?

<!-- gh-comment-id:2304467810 --> @NightingaleCen commented on GitHub (Aug 22, 2024): > When I use "ollama run phi3.5", the model replies but it does not stop. It just continues with anything else until the token window is full. Might be related to ggerganov/llama.cpp#9127 ?
GiteaMirror commented 2026-05-03 23:57:23 -05:00
Author
Owner
Copy Link

@badvision commented on GitHub (Aug 22, 2024):

Using "ollama run phi3.5" works for a short while and then the output turns completely erratic.

You can start the conversation simply asking "Did you understand the question I just asked?" and it will start generating nonsense.

<!-- gh-comment-id:2305356114 --> @badvision commented on GitHub (Aug 22, 2024): > Using "ollama run phi3.5" works for a short while and then the output turns completely erratic. You can start the conversation simply asking "Did you understand the question I just asked?" and it will start generating nonsense.
GiteaMirror commented 2026-05-03 23:57:24 -05:00
Author
Owner
Copy Link

@terodactilo21 commented on GitHub (Aug 22, 2024):

Vision model could be awesome.

<!-- gh-comment-id:2305763716 --> @terodactilo21 commented on GitHub (Aug 22, 2024): Vision model could be awesome.
GiteaMirror commented 2026-05-03 23:57:26 -05:00
Author
Owner
Copy Link

@daboe01 commented on GitHub (Aug 31, 2024):

https://github.com/ggerganov/llama.cpp/pull/9209

<!-- gh-comment-id:2322921745 --> @daboe01 commented on GitHub (Aug 31, 2024): https://github.com/ggerganov/llama.cpp/pull/9209
GiteaMirror commented 2026-05-03 23:57:27 -05:00
Author
Owner
Copy Link

@zcasanova commented on GitHub (Sep 6, 2024):

I am also seeing this behaviour

<!-- gh-comment-id:2334559832 --> @zcasanova commented on GitHub (Sep 6, 2024): I am also seeing this behaviour
GiteaMirror commented 2026-05-03 23:57:28 -05:00
Author
Owner
Copy Link

@smashweaver commented on GitHub (Sep 14, 2024):

ollama run phi3.5
pulling manifest
pulling b5374915da53... 100% ▕██████████████████████████████████████████████████████████████████▏ 2.2 GB
pulling c608dc615584... 100% ▕██████████████████████████████████████████████████████████████████▏ 149 B
pulling c5f9f560a704... 100% ▕██████████████████████████████████████████████████████████████████▏ 1.1 KB
pulling 45ac4c662a77... 100% ▕██████████████████████████████████████████████████████████████████▏ 60 B
pulling 0611867ea67b... 100% ▕██████████████████████████████████████████████████████████████████▏ 409 B
verifying sha256 digest
Error: digest mismatch, file must be downloaded again: want sha256:b5374915da534cb93df39f03bd4f2cd5a0c533df0d5e21957dc9556c260be9eb, got sha256:ea6cbaac56d2799594481b797f0e3a2c99a4fef7fba5aab7f0afa7827cb1dce7

<!-- gh-comment-id:2350747460 --> @smashweaver commented on GitHub (Sep 14, 2024): ollama run phi3.5 pulling manifest pulling b5374915da53... 100% ▕██████████████████████████████████████████████████████████████████▏ 2.2 GB pulling c608dc615584... 100% ▕██████████████████████████████████████████████████████████████████▏ 149 B pulling c5f9f560a704... 100% ▕██████████████████████████████████████████████████████████████████▏ 1.1 KB pulling 45ac4c662a77... 100% ▕██████████████████████████████████████████████████████████████████▏ 60 B pulling 0611867ea67b... 100% ▕██████████████████████████████████████████████████████████████████▏ 409 B verifying sha256 digest Error: digest mismatch, file must be downloaded again: want sha256:b5374915da534cb93df39f03bd4f2cd5a0c533df0d5e21957dc9556c260be9eb, got sha256:ea6cbaac56d2799594481b797f0e3a2c99a4fef7fba5aab7f0afa7827cb1dce7
GiteaMirror commented 2026-05-03 23:57:28 -05:00
Author
Owner
Copy Link

@gaussiangit commented on GitHub (Nov 15, 2024):

Is it possible to use Phi-3.5-vision-instruct ? I believe onnx versions are available but not sure how to port to ollama ?

<!-- gh-comment-id:2478774727 --> @gaussiangit commented on GitHub (Nov 15, 2024): Is it possible to use [Phi-3.5-vision-instruct](https://huggingface.co/microsoft/Phi-3.5-vision-instruct) ? I believe onnx versions are available but not sure how to port to ollama ?
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 model
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#66094
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?