Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

OLLAMA models support? v1.1.0 #135

Open
saibalz2ai opened this issue Nov 4, 2024 · 14 comments
Open

OLLAMA models support? v1.1.0 #135

saibalz2ai opened this issue Nov 4, 2024 · 14 comments

Comments

@saibalz2ai
Copy link

Which local model supports running UFO with ollama? I have tried with the latest models, but not a single one worked.
python -m ufo --task ollama1

Welcome to use UFO🛸, A UI-focused Agent for Windows OS Interaction.

| | | || | / _
| | | || | | | | |
| || || | | || |
_/ |_| __/

Please enter your request to be completed🛸:
open word, bold word breakup
Round 1, Step 1, HostAgent: Analyzing the user intent and decomposing the request...
Observations👀: None
Thoughts💡: None
Plans📚: (1) None
(2)
Next Selected application📲: [The required application needs to be opened.]
Messages to AppAgent📩:
Status📊: None
Comment💬: None
Creating an experience indexer...
Warning: Failed to load experience indexer from vectordb/experience/experience_db.
Creating an demonstration indexer...
Warning: Failed to load demonstration indexer from vectordb/demonstration/demonstration_db.
Please enter your new request. Enter 'N' for exit.
@saibalz2ai saibalz2ai changed the title OLLAMA models support? OLLAMA models support? v1.1.0 Nov 4, 2024
@showkeyjar
Copy link

+1

@vyokky
Copy link
Contributor

vyokky commented Nov 23, 2024

@saibalz2ai @showkeyjar it is supported, please check this document: https://microsoft.github.io/UFO/supported_models/ollama/

@saibalz2ai
Copy link
Author

saibalz2ai commented Nov 23, 2024 via email

@linmiao5
Copy link

linmiao5 commented Dec 2, 2024 •
edited
Loading

修改了 API_TYPE: "Ollama" ；API_BASE: "http://localhost:11434"；API_MODEL: "llava-llama3"，出现了Creating an experience indexer...
Warning: Failed to load experience indexer from vectordb/experience/experience_db.
Creating an demonstration indexer...
Warning: Failed to load demonstration indexer from vectordb/demonstration/demonstration_db.如何解决？

@vyokky
Copy link
Contributor

vyokky commented Dec 5, 2024

修改了 API_TYPE: "Ollama" ；API_BASE: "http://localhost:11434"；API_MODEL: "llava-llama3"，出现了Creating an experience indexer... Warning: Failed to load experience indexer from vectordb/experience/experience_db. Creating an demonstration indexer... Warning: Failed to load demonstration indexer from vectordb/demonstration/demonstration_db.如何解决？

@linmiao5 This is only a warning, which will affect the main workflow. Do you encounter other issues?

@linmiao5
Copy link

linmiao5 commented Dec 5, 2024

Which Ollama model did you use to run this project successfully?

@Mac0q
Copy link
Contributor

Mac0q commented Dec 5, 2024

@linmiao5 At the beginning, llava:7b was used to have a taste on UFO, but due to its own context length limitation, it needs to satisfy the input restriction of llava itself by removing example and other prompts. It is recommended to choose a model that supports a long context to have a try, of course GPT is for sure the best choice.

@icebilly
Copy link

icebilly commented Feb 6, 2025

I always get the connection timeout error when using a localhosted ollama model.

Image

@saibalz2ai
Copy link
Author

saibalz2ai commented Feb 6, 2025 via email

@icebilly
Copy link

icebilly commented Feb 7, 2025

Check ollama access through curl command from the machine you are trying to connect.
You will get curl command example from ollama page.

Get Outlook for Androidhttps://aka.ms/AAb9ysg
…

I extend the timeout to 3 mins, the timeout issue gone, but still have the issue you mentioned in the first post.

@icebilly
Copy link

icebilly commented Feb 7, 2025

I use deepseek-r1:14b served with ollama by the way.

@saibalz2ai
Copy link
Author

saibalz2ai commented Feb 7, 2025 via email

@icebilly
Copy link

icebilly commented Feb 7, 2025

That doesn't matter with timeout. Host timeout means ollama is not reachable with the given ip:port.
I have provided you the example location. Try that.
Reg. Deepseek I am not sure ufo supporting this model or not.
1st you resolved the timeout issue.

On 07-Feb-2025 6:48 am, icebilly @.***> wrote:

I use deepseek-r1:14b served with ollama by the way.

—
Reply to this email directly, view it on GitHubhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_microsoft_UFO_issues_135-23issuecomment-2D2641595385&d=DwMCaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=8kZK3igKo8ofL0_RJ4VxHNDpyW3p9Ly_jyMqmaBoeHY&m=B1wYuU2dPv-yfWLlg2cnJ98oWT2uOtZcHxCBQO9EG4qcna_cE7HVOT_pJ0dYOxfC&s=F3ptmuiMwB7NniPVJKJgTA14ov3vB347eShMqKWtHnA&e=, or unsubscribehttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AZGVKCDPFOI4Z6QYM57WMQ32OQCWRAVCNFSM6AAAAABRDVBHKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMNBRGU4TKMZYGU&d=DwMCaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=8kZK3igKo8ofL0_RJ4VxHNDpyW3p9Ly_jyMqmaBoeHY&m=B1wYuU2dPv-yfWLlg2cnJ98oWT2uOtZcHxCBQO9EG4qcna_cE7HVOT_pJ0dYOxfC&s=YtwUmbSl6DfgjUjXgggJZu5Jp5rvloEvnZHVSzkjG5Q&e=.
You are receiving this because you were mentioned.Message ID: @.***>

the curl command seems ok, below is what I get when using deepseek-r1:14b as the host agent and app agent.

Image

@saibalz2ai
Copy link
Author

saibalz2ai commented Feb 7, 2025 via email

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
6 participants
@vyokky @showkeyjar @icebilly @Mac0q @saibalz2ai @linmiao5