Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-19 14:13:35 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #3042] [BUG]: Agent stuck #1947

New Issue
Closed
opened 2026-02-22 18:27:22 -05:00 by yindo · 3 comments
yindo commented 2026-02-22 18:27:22 -05:00
Owner
Copy Link
Copy Source

Originally created by @Namelesssguy on GitHub (Jan 27, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/3042

How are you running AnythingLLM?

AnythingLLM desktop app

What happened?

Image
Agent does not do anything after showing -- Agent @agent invoked swapping over to agent chat. Type /exit to exit execution loop early. I have tried it with deepseek r1 8 billion parameter model, and llama 8 billion parameter model in both case it take about 20 minutes to show agent session complete after giving the responce.

Image

Are there known steps to reproduce?

No response

Originally created by @Namelesssguy on GitHub (Jan 27, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/3042 ### How are you running AnythingLLM? AnythingLLM desktop app ### What happened? ![Image](https://github.com/user-attachments/assets/d1943d9a-49e6-4994-a66a-4939217bacbf) Agent does not do anything after showing -- Agent @agent invoked swapping over to agent chat. Type /exit to exit execution loop early. I have tried it with deepseek r1 8 billion parameter model, and llama 8 billion parameter model in both case it take about 20 minutes to show agent session complete after giving the responce. ![Image](https://github.com/user-attachments/assets/03ade1cb-ecce-4fdc-b5bd-dbbafe4c6583) ### Are there known steps to reproduce? _No response_
yindo added the possible bug label 2026-02-22 18:27:22 -05:00
yindo closed this issue 2026-02-22 18:27:22 -05:00
yindo commented 2026-02-22 18:27:22 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Jan 28, 2025):

The performance of agents is 100% proportional to your machine specs being able to run model inference quickly. its the same as if there was a really large prompt or document in the context window.

You didn't mention the specs of your machine, but I am going to assume you are running this on CPU only if it is taking that long to get a response.

Do you have a GPU, what are your CPU and RAM specs?

@timothycarambat commented on GitHub (Jan 28, 2025): The performance of agents is 100% proportional to your machine specs being able to run model inference quickly. its the same as if there was a really large prompt or document in the context window. You didn't mention the specs of your machine, but I am going to assume you are running this on CPU only if it is taking that long to get a response. Do you have a GPU, what are your CPU and RAM specs?
yindo commented 2026-02-22 18:27:22 -05:00
Author
Owner
Copy Link
Copy Source

@Namelesssguy commented on GitHub (Jan 28, 2025):

The performance of agents is 100% proportional to your machine specs being able to run model inference quickly. its the same as if there was a really large prompt or document in the context window.

You didn't mention the specs of your machine, but I am going to assume you are running this on CPU only if it is taking that long to get a response.

Do you have a GPU, what are your CPU and RAM specs?

My specs are intel i7 10750H and gpu rtx 3050 4GB, and 16GB ddr4 RAM. Also I don't think you understood my problem correctly, the responce from the llm is instant, its the agent session that take time, if i wait for 20mins, then it shows (agent session complete).

@Namelesssguy commented on GitHub (Jan 28, 2025): > The performance of agents is 100% proportional to your machine specs being able to run model inference quickly. its the same as if there was a really large prompt or document in the context window. > > You didn't mention the specs of your machine, but I am going to assume you are running this on CPU only if it is taking that long to get a response. > > Do you have a GPU, what are your CPU and RAM specs? My specs are intel i7 10750H and gpu rtx 3050 4GB, and 16GB ddr4 RAM. Also I don't think you understood my problem correctly, the responce from the llm is instant, its the agent session that take time, if i wait for 20mins, then it shows (agent session complete).
yindo commented 2026-02-22 18:27:22 -05:00
Author
Owner
Copy Link
Copy Source

@timothycarambat commented on GitHub (Jan 28, 2025):

if i wait for 20mins, then it shows (agent session complete).

This is intended, we auto-close the websocket for agent handler if nothing is sent after a period of time from the client. A "session" is not a single question, a session is being in agent mode.

So you would send a query, get a response like you did, and then you could keep chatting while the LLM is now using tools. You don't need to wait for the session to end, you just need to wait for a response.

You can have a whole chat going back and forth, using tools, while in a single-agent session. Closing of an agent session just means the chat is now going back to RAG-chat only mode.

Most people open a session and just use AnythingLLM in agent mode when they want a continuous conversation with tool-calling to be possible.

@timothycarambat commented on GitHub (Jan 28, 2025): > if i wait for 20mins, then it shows (agent session complete). This is intended, we auto-close the websocket for agent handler if nothing is sent after a period of time from the client. A "session" is not a single question, a session is being in agent mode. So you would send a query, get a response like you did, and then you could keep chatting while the LLM is now using tools. You don't need to wait for the session to end, you just need to wait for a response. You can have a whole chat going back and forth, using tools, while in a single-agent session. Closing of an agent session just means the chat is now going back to RAG-chat only mode. Most people open a session and just use AnythingLLM in agent mode when they want a continuous conversation with tool-calling to be possible.
yindo changed title from [BUG]: Agent stuck to [GH-ISSUE #3042] [BUG]: Agent stuck 2026-06-05 14:43:37 -04:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label possible bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#1947
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?