github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 13:19:40 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #1574] Sending several requests to the server in quick succession appears to cause some responses to fail #26624

New Issue
Closed
opened 2026-04-22 03:00:26 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-22 03:00:26 -05:00
Owner
Copy Link

Originally created by @charstorm on GitHub (Dec 17, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/1574

Hi,

First, I want to thank everyone working on this project. I appreciate your efforts.

I was testing ollama server and I noticed that it sometimes gave empty responses. I found out that it happens when a request is made right after the previous one. Adding a sleep seems to solve the issue. Here is some code to demonstrate the issue:


import time
import json
import requests

DEFAULT_URL = "http://localhost:11434/api/generate"
DEFAULT_MODEL = "mistral"

def generate(prompt, model=DEFAULT_MODEL, url=DEFAULT_URL):
    post_data = {
        "prompt": prompt, "model": model, "stream": True
    }
    response = requests.post(url, json=post_data)
    response.raise_for_status()
    parts = []
    for line in response.iter_lines():
        body = json.loads(line)
        if "error" in body:
            raise Exception(body["error"])

        content = body.get("response", "")
        parts.append(content)

        if body.get("done"):
            break

    return "".join(parts).strip()


def test():
    prompts = [
        "What is radian?", "What is meridian?",
        "What is steradian?", "What is circadian?"
    ]
    for prompt in prompts:
        print("Prompt:", prompt)
        response = generate(prompt)
        if not response:
            print("ERROR: got empty response!")
        else:
            print("Response:", response)
        print("----"*20)
        # time.sleep(0.3)  # <-- this is needed to avoid empty responses


if __name__ == "__main__":
    test()
Originally created by @charstorm on GitHub (Dec 17, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/1574 Hi, First, I want to thank everyone working on this project. I appreciate your efforts. I was testing ollama server and I noticed that it sometimes gave empty responses. I found out that it happens when a request is made right after the previous one. Adding a sleep seems to solve the issue. Here is some code to demonstrate the issue: ```python import time import json import requests DEFAULT_URL = "http://localhost:11434/api/generate" DEFAULT_MODEL = "mistral" def generate(prompt, model=DEFAULT_MODEL, url=DEFAULT_URL): post_data = { "prompt": prompt, "model": model, "stream": True } response = requests.post(url, json=post_data) response.raise_for_status() parts = [] for line in response.iter_lines(): body = json.loads(line) if "error" in body: raise Exception(body["error"]) content = body.get("response", "") parts.append(content) if body.get("done"): break return "".join(parts).strip() def test(): prompts = [ "What is radian?", "What is meridian?", "What is steradian?", "What is circadian?" ] for prompt in prompts: print("Prompt:", prompt) response = generate(prompt) if not response: print("ERROR: got empty response!") else: print("Response:", response) print("----"*20) # time.sleep(0.3) # <-- this is needed to avoid empty responses if __name__ == "__main__": test() ```
GiteaMirror commented 2026-04-22 03:00:28 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Dec 17, 2023):

Hi @charstorm, thanks for creating an issue (and including some code to reproduce it!)

This issue should have been fixed as of 0.1.14: https://github.com/jmorganca/ollama/releases/tag/v0.1.14

Do you know which version of Ollama you might be running (you can check with ollama -v)? If it's below 0.1.14, try upgrading and let me know if that fixes it!

<!-- gh-comment-id:1859264252 --> @jmorganca commented on GitHub (Dec 17, 2023): Hi @charstorm, thanks for creating an issue (and including some code to reproduce it!) This issue should have been fixed as of 0.1.14: https://github.com/jmorganca/ollama/releases/tag/v0.1.14 Do you know which version of Ollama you might be running (you can check with `ollama -v`)? If it's below 0.1.14, try upgrading and let me know if that fixes it!
GiteaMirror commented 2026-04-22 03:00:29 -05:00
Author
Owner
Copy Link

@charstorm commented on GitHub (Dec 17, 2023):

You are right. I was on 0.1.13. Just downloaded and tested 0.1.14 and it works without having to add sleep.
Thank you!

<!-- gh-comment-id:1859268000 --> @charstorm commented on GitHub (Dec 17, 2023): You are right. I was on 0.1.13. Just downloaded and tested 0.1.14 and it works without having to add sleep. Thank you!
GiteaMirror commented 2026-04-22 03:00:29 -05:00
Author
Owner
Copy Link

@igorschlum commented on GitHub (Dec 17, 2023):

@charstorm Great. Thank you for the script. Could you please mark the issue as Closed ?

<!-- gh-comment-id:1859271109 --> @igorschlum commented on GitHub (Dec 17, 2023): @charstorm Great. Thank you for the script. Could you please mark the issue as Closed ?
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
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#26624
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?