github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #9711] Error: POST Predict Request to http://127.0.0.1:53151/completion Failed – Connection Forcibly Closed by Remote Host (wsarecv) #6344

New Issue
Closed
opened 2026-04-12 17:51:25 -05:00 by GiteaMirror · 15 comments
GiteaMirror commented 2026-04-12 17:51:25 -05:00
Owner
Copy Link

Originally created by @Dejon141 on GitHub (Mar 13, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9711

What is the issue?

So I was running the new gemma3 models when I stumbled upon a error with gemma3:27b didn't know why it happened even after testing multiple times also I don't have any environment variables interfering neither do I have an external host accessing the server tried it with the different gemma models and it worked it seems like either my GPU is a potato or there is some kinda of internal error with the ollama server.

Error: POST predict: Post "http://127.0.0.1:53194/completion": read tcp 127.0.0.1:53196->127.0.0.1:53194: wsarecv: An existing connection was forcibly closed by the remote host.

COMMAND

"ollama run gemma3:27b"

SPECIFICATION

Geforce GTX 1060 6GB
Intel i5 10400F

Relevant log output

[server.log](https://github.com/user-attachments/files/19222811/server.log)

OS

Windows

GPU

Nvidia

CPU

Intel

Ollama version

0.6.0

Originally created by @Dejon141 on GitHub (Mar 13, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9711 ### What is the issue? So I was running the new gemma3 models when I stumbled upon a error with gemma3:27b didn't know why it happened even after testing multiple times also I don't have any environment variables interfering neither do I have an external host accessing the server tried it with the different gemma models and it worked it seems like either my GPU is a potato or there is some kinda of internal error with the ollama server. Error: POST predict: Post "http://127.0.0.1:53194/completion": read tcp 127.0.0.1:53196->127.0.0.1:53194: wsarecv: An existing connection was forcibly closed by the remote host. ### COMMAND "ollama run gemma3:27b" ## SPECIFICATION Geforce GTX 1060 6GB Intel i5 10400F ### Relevant log output ```shell [server.log](https://github.com/user-attachments/files/19222811/server.log) ``` ### OS Windows ### GPU Nvidia ### CPU Intel ### Ollama version 0.6.0
GiteaMirror added the bug label 2026-04-12 17:51:25 -05:00
GiteaMirror commented 2026-04-12 17:51:28 -05:00
Author
Owner
Copy Link

@sabre-code commented on GitHub (Mar 13, 2025):

same here

ResponseError: POST predict: Post "http://127.0.0.1:40133/completion": EOF

<!-- gh-comment-id:2720092624 --> @sabre-code commented on GitHub (Mar 13, 2025): same here ResponseError: POST predict: Post "http://127.0.0.1:40133/completion": EOF
GiteaMirror commented 2026-04-12 17:51:29 -05:00
Author
Owner
Copy Link

@vlimki commented on GitHub (Mar 13, 2025):

Same here.

<!-- gh-comment-id:2720197627 --> @vlimki commented on GitHub (Mar 13, 2025): Same here.
GiteaMirror commented 2026-04-12 17:51:29 -05:00
Author
Owner
Copy Link

@stonega commented on GitHub (Mar 13, 2025):

Same hare

<!-- gh-comment-id:2720213217 --> @stonega commented on GitHub (Mar 13, 2025): Same hare
GiteaMirror commented 2026-04-12 17:51:30 -05:00
Author
Owner
Copy Link

@titandino88 commented on GitHub (Mar 13, 2025):

same here

<!-- gh-comment-id:2720353270 --> @titandino88 commented on GitHub (Mar 13, 2025): same here
GiteaMirror commented 2026-04-12 17:51:30 -05:00
Author
Owner
Copy Link

@SimpleYj commented on GitHub (Mar 13, 2025):

same here:Error: POST predict: Post "http://127.0.0.1:43845/completion": EOF

<!-- gh-comment-id:2720370200 --> @SimpleYj commented on GitHub (Mar 13, 2025): same here:Error: POST predict: Post "http://127.0.0.1:43845/completion": EOF
GiteaMirror commented 2026-04-12 17:51:31 -05:00
Author
Owner
Copy Link

@vlimki commented on GitHub (Mar 13, 2025):

https://github.com/ollama/ollama/issues/9699#issuecomment-2719330180

This helped me out

<!-- gh-comment-id:2720445003 --> @vlimki commented on GitHub (Mar 13, 2025): https://github.com/ollama/ollama/issues/9699#issuecomment-2719330180 This helped me out
GiteaMirror commented 2026-04-12 17:51:31 -05:00
Author
Owner
Copy Link

@ACheshirov commented on GitHub (Mar 13, 2025):

Same here: Error: POST predict: Post "http://127.0.0.1:43569/completion": EOF

<!-- gh-comment-id:2720639817 --> @ACheshirov commented on GitHub (Mar 13, 2025): Same here: Error: POST predict: Post "http://127.0.0.1:43569/completion": EOF
GiteaMirror commented 2026-04-12 17:51:31 -05:00
Author
Owner
Copy Link

@tuber84 commented on GitHub (Mar 13, 2025):

Same problem

<!-- gh-comment-id:2721364993 --> @tuber84 commented on GitHub (Mar 13, 2025): Same problem
GiteaMirror commented 2026-04-12 17:51:32 -05:00
Author
Owner
Copy Link

@omargohan commented on GitHub (Mar 14, 2025):

I had the same issue until I tried adding Environment="GGML_CUDA_ENABLE_UNIFIED_MEMORY=1" to ollama.service as suggested here.

<!-- gh-comment-id:2723816084 --> @omargohan commented on GitHub (Mar 14, 2025): I had the same issue until I tried adding `Environment="GGML_CUDA_ENABLE_UNIFIED_MEMORY=1"` to `ollama.service` as suggested [here](https://github.com/ollama/ollama/issues/9707#issuecomment-2719746522).
GiteaMirror commented 2026-04-12 17:51:32 -05:00
Author
Owner
Copy Link

@Walker555 commented on GitHub (Mar 14, 2025):

Fix for Docker users:
docker run -d --gpus=all -v /path/to/my/volumes/ollama:/root/.ollama -p 11434:11434 -e GGML_CUDA_ENABLE_UNIFIED_MEMORY=1 --name ollama ollama/ollama:0.6.0

<!-- gh-comment-id:2723915202 --> @Walker555 commented on GitHub (Mar 14, 2025): Fix for Docker users: `docker run -d --gpus=all -v /path/to/my/volumes/ollama:/root/.ollama -p 11434:11434 -e GGML_CUDA_ENABLE_UNIFIED_MEMORY=1 --name ollama ollama/ollama:0.6.0`
GiteaMirror commented 2026-04-12 17:51:33 -05:00
Author
Owner
Copy Link

@openelearning commented on GitHub (Mar 14, 2025):

Same problem with or without Environment="GGML_CUDA_ENABLE_UNIFIED_MEMORY=1" with gemma 27b or 12b, and with nvidia cuda.

<!-- gh-comment-id:2724892810 --> @openelearning commented on GitHub (Mar 14, 2025): Same problem with or without Environment="GGML_CUDA_ENABLE_UNIFIED_MEMORY=1" with gemma 27b or 12b, and with nvidia cuda.
GiteaMirror commented 2026-04-12 17:51:35 -05:00
Author
Owner
Copy Link

@kurokirasama commented on GitHub (Mar 15, 2025):

just reporting that the problem is not fixed in 0.6.1 for gemma3:27b in my main machine, but it was fixed in another machine (with lower resources), with the error in gemma3:4b.

<!-- gh-comment-id:2726982260 --> @kurokirasama commented on GitHub (Mar 15, 2025): just reporting that the problem is not fixed in 0.6.1 for gemma3:27b in my main machine, but it was fixed in another machine (with lower resources), with the error in gemma3:4b.
GiteaMirror commented 2026-04-12 17:51:36 -05:00
Author
Owner
Copy Link

@fansRealx commented on GitHub (Mar 24, 2025):

Same problem ，The number of words in your question exceeds the maximum token limit of the model. You can set the model's token length longer.

<!-- gh-comment-id:2746788713 --> @fansRealx commented on GitHub (Mar 24, 2025): Same problem ，The number of words in your question exceeds the maximum token limit of the model. You can set the model's token length longer.
GiteaMirror commented 2026-04-12 17:51:37 -05:00
Author
Owner
Copy Link

@tunisiano187 commented on GitHub (May 1, 2025):

0.6.6 has the same error, back to 0.6.5, solved it

<!-- gh-comment-id:2845262311 --> @tunisiano187 commented on GitHub (May 1, 2025): 0.6.6 has the same error, back to 0.6.5, solved it
GiteaMirror commented 2026-04-12 17:51:39 -05:00
Author
Owner
Copy Link

@jessegross commented on GitHub (May 1, 2025):

I believe that the original issue has been resolved, so I'm going to close this. If you are still seeing this with 0.6.7, please create a new issue and attach logs.

<!-- gh-comment-id:2845688461 --> @jessegross commented on GitHub (May 1, 2025): I believe that the original issue has been resolved, so I'm going to close this. If you are still seeing this with 0.6.7, please create a new issue and attach logs.
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 bug
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#6344
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?