Bug: Text Generation Process Fails Silently in Ollama Serve #201

Closed
opened 2025-11-11 14:11:11 -06:00 by GiteaMirror · 1 comment
Owner

Originally created by @hemangjoshi37a on GitHub (Jan 18, 2024).

Issue Description:
At times, when I use the "generate" button to initiate text generation after entering a prompt, the process fails to produce any output. This failure occurs without any error messages. Observing the htop terminal, it appears that the ollama serve process either stops or completes. However, the chat window continues to display the shimmer effect, which usually indicates ongoing generation, leading to confusion about the actual state of the process.

Steps to Reproduce:

  1. Enter a text prompt in the chat window.
  2. Click on the "generate" button to start the text generation process.
  3. Observe the lack of output in the chat window and the status of the process in the htop terminal.

Expected Behavior:
When the text generation process is initiated, it should either successfully generate text or provide an error message if it fails.

Actual Behavior:
The text generation process stops or completes silently without any output in the chat window, while the ollama serve process appears to stop in the htop terminal. The chat window continues to show the shimmer effect, incorrectly suggesting that generation is still in progress.

Environment:

  • python 3.11 and latest ollama version installed yesterday using github readme.
Originally created by @hemangjoshi37a on GitHub (Jan 18, 2024). **Issue Description:** At times, when I use the "generate" button to initiate text generation after entering a prompt, the process fails to produce any output. This failure occurs without any error messages. Observing the `htop` terminal, it appears that the `ollama serve` process either stops or completes. However, the chat window continues to display the shimmer effect, which usually indicates ongoing generation, leading to confusion about the actual state of the process. **Steps to Reproduce:** 1. Enter a text prompt in the chat window. 2. Click on the "generate" button to start the text generation process. 3. Observe the lack of output in the chat window and the status of the process in the `htop` terminal. **Expected Behavior:** When the text generation process is initiated, it should either successfully generate text or provide an error message if it fails. **Actual Behavior:** The text generation process stops or completes silently without any output in the chat window, while the `ollama serve` process appears to stop in the `htop` terminal. The chat window continues to show the shimmer effect, incorrectly suggesting that generation is still in progress. **Environment:** - python 3.11 and latest ollama version installed yesterday using github readme.
Author
Owner

@tjbck commented on GitHub (Jan 18, 2024):

cannot reproduce the issue, moving to discussion.

@tjbck commented on GitHub (Jan 18, 2024): cannot reproduce the issue, moving to discussion.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#201