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 #8427] 0.5.5 Model Creation on Windows Seems Broken #51927

New Issue
Closed
opened 2026-04-28 21:16:08 -05:00 by GiteaMirror · 13 comments
GiteaMirror commented 2026-04-28 21:16:08 -05:00
Owner
Copy Link

Originally created by @github-pmj on GitHub (Jan 14, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/8427

Originally assigned to: @pdevine on GitHub.

What is the issue?

Version 0.5.5 seems to have broken create functionality on Windows. I use a PowerShell script to create new models with different templates. The first step in the script is to run
ollama create '<model name>' -f <full path to model file>

The model file is a one line file
FROM <full path to GGUF>

With 0.5.5 it errors on model creation as it mangles the file path (seems to pre-pend the OLLAMA_MODELS environment variable to the GGUF path). When I modify the model file to take this into account and run
ollama create 'testmodel1' -f <full path to model file>
directly I get an 'Invalid model name' error.

Rolling back to 0.5.4 resolves the issue.

I'm using the Ollama setup installer.

OS

Windows

GPU

Nvidia

CPU

AMD

Ollama version

0.5.5

Originally created by @github-pmj on GitHub (Jan 14, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/8427 Originally assigned to: @pdevine on GitHub. ### What is the issue? Version 0.5.5 seems to have broken create functionality on Windows. I use a PowerShell script to create new models with different templates. The first step in the script is to run ` ollama create '<model name>' -f <full path to model file>` The model file is a one line file ` FROM <full path to GGUF>` With 0.5.5 it errors on model creation as it mangles the file path (seems to pre-pend the OLLAMA_MODELS environment variable to the GGUF path). When I modify the model file to take this into account and run ` ollama create 'testmodel1' -f <full path to model file>` directly I get an 'Invalid model name' error. Rolling back to 0.5.4 resolves the issue. I'm using the Ollama setup installer. ### OS Windows ### GPU Nvidia ### CPU AMD ### Ollama version 0.5.5
GiteaMirror added the bug label 2026-04-28 21:16:09 -05:00
GiteaMirror commented 2026-04-28 21:16:11 -05:00
Author
Owner
Copy Link

@iSevenDays commented on GitHub (Jan 14, 2025):

I can confirm on Linux it is broken too

(base) root@725870ad64ad:~# cat /tmp/modelfile 
# Modelfile generated by "ollama show"
# To build a new Modelfile based on this, replace FROM with:
# FROM hf.co/bartowski/glm-4-9b-chat-1m-GGUF:Q5_K_M

FROM /root/.ollama/models/blobs/sha256-3c874d569aef8e0b7be7d46aa6bc1d274630eafaaa65f54107151be4ef868e7e
PARAMETER num_ctx 409600
TEMPLATE "[gMASK]<sop>{{ if .System }}<|system|>
{{ .System }}{{ end }}{{ if .Prompt }}<|user|>
{{ .Prompt }}{{ end }}<|assistant|>
{{ .Response }}"
PARAMETER stop <|system|>
PARAMETER stop <|user|>
PARAMETER stop <|assistant|>

(base) root@725870ad64ad:~# ollama create example -f /tmp/modelfile 
gathering model components 
Error: invalid model name
(base) root@725870ad64ad:~# echo $MODEL_NAME
hf.co/bartowski/glm-4-9b-chat-1m-GGUF:Q5_K_M
(base) root@725870ad64ad:~# ollama show "$MODEL_NAME" --modelfile > /tmp/modelfile
(base) root@725870ad64ad:~# cat /tmp/modelfile 
# Modelfile generated by "ollama show"
# To build a new Modelfile based on this, replace FROM with:
# FROM hf.co/bartowski/glm-4-9b-chat-1m-GGUF:Q5_K_M

FROM /root/.ollama/models/blobs/sha256-3c874d569aef8e0b7be7d46aa6bc1d274630eafaaa65f54107151be4ef868e7e
TEMPLATE "[gMASK]<sop>{{ if .System }}<|system|>
{{ .System }}{{ end }}{{ if .Prompt }}<|user|>
{{ .Prompt }}{{ end }}<|assistant|>
{{ .Response }}"
PARAMETER stop <|system|>
PARAMETER stop <|user|>
PARAMETER stop <|assistant|>

(base) root@725870ad64ad:~# ollama create example -f /tmp/modelfile 
gathering model components 
Error: invalid model name
(base) root@725870ad64ad:~#
<!-- gh-comment-id:2591185444 --> @iSevenDays commented on GitHub (Jan 14, 2025): I can confirm on Linux it is broken too ``` (base) root@725870ad64ad:~# cat /tmp/modelfile # Modelfile generated by "ollama show" # To build a new Modelfile based on this, replace FROM with: # FROM hf.co/bartowski/glm-4-9b-chat-1m-GGUF:Q5_K_M FROM /root/.ollama/models/blobs/sha256-3c874d569aef8e0b7be7d46aa6bc1d274630eafaaa65f54107151be4ef868e7e PARAMETER num_ctx 409600 TEMPLATE "[gMASK]<sop>{{ if .System }}<|system|> {{ .System }}{{ end }}{{ if .Prompt }}<|user|> {{ .Prompt }}{{ end }}<|assistant|> {{ .Response }}" PARAMETER stop <|system|> PARAMETER stop <|user|> PARAMETER stop <|assistant|> (base) root@725870ad64ad:~# ollama create example -f /tmp/modelfile gathering model components Error: invalid model name (base) root@725870ad64ad:~# echo $MODEL_NAME hf.co/bartowski/glm-4-9b-chat-1m-GGUF:Q5_K_M (base) root@725870ad64ad:~# ollama show "$MODEL_NAME" --modelfile > /tmp/modelfile (base) root@725870ad64ad:~# cat /tmp/modelfile # Modelfile generated by "ollama show" # To build a new Modelfile based on this, replace FROM with: # FROM hf.co/bartowski/glm-4-9b-chat-1m-GGUF:Q5_K_M FROM /root/.ollama/models/blobs/sha256-3c874d569aef8e0b7be7d46aa6bc1d274630eafaaa65f54107151be4ef868e7e TEMPLATE "[gMASK]<sop>{{ if .System }}<|system|> {{ .System }}{{ end }}{{ if .Prompt }}<|user|> {{ .Prompt }}{{ end }}<|assistant|> {{ .Response }}" PARAMETER stop <|system|> PARAMETER stop <|user|> PARAMETER stop <|assistant|> (base) root@725870ad64ad:~# ollama create example -f /tmp/modelfile gathering model components Error: invalid model name (base) root@725870ad64ad:~# ```
GiteaMirror commented 2026-04-28 21:16:13 -05:00
Author
Owner
Copy Link

@chllei commented on GitHub (Jan 15, 2025):

The NVIDIA Jetson gets the same issue.

<!-- gh-comment-id:2593641483 --> @chllei commented on GitHub (Jan 15, 2025): The NVIDIA Jetson gets the same issue.
GiteaMirror commented 2026-04-28 21:16:14 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (Jan 15, 2025):

@chllei Please try 0.5.6. It should be fixed.

<!-- gh-comment-id:2594169915 --> @pdevine commented on GitHub (Jan 15, 2025): @chllei Please try 0.5.6. It should be fixed.
GiteaMirror commented 2026-04-28 21:16:16 -05:00
Author
Owner
Copy Link

@lilyrosecarnation commented on GitHub (Jan 20, 2025):

@chllei Please try 0.5.6. It should be fixed.

Not yet. Still yield the error.

<!-- gh-comment-id:2601830221 --> @lilyrosecarnation commented on GitHub (Jan 20, 2025): > [@chllei](https://github.com/chllei) Please try 0.5.6. It should be fixed. Not yet. Still yield the error.
GiteaMirror commented 2026-04-28 21:16:17 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (Jan 21, 2025):

Just make certain both the client and the server are up-to-date. Can you do ollama -v?

<!-- gh-comment-id:2605680937 --> @pdevine commented on GitHub (Jan 21, 2025): Just make certain both the client and the server are up-to-date. Can you do `ollama -v`?
GiteaMirror commented 2026-04-28 21:16:21 -05:00
Author
Owner
Copy Link

@Sherlock-Holo commented on GitHub (Feb 3, 2025):

problem still exists, I am running ollama on Windows

ollama create DeepSeek-R1-Distill-Llama-8B -f .\Modelfile
gathering model components
Error: invalid model name

ollama -v

ollama version is 0.5.7

<!-- gh-comment-id:2630602385 --> @Sherlock-Holo commented on GitHub (Feb 3, 2025): problem still exists, I am running ollama on Windows > ollama create DeepSeek-R1-Distill-Llama-8B -f .\Modelfile > gathering model components > Error: invalid model name `ollama -v` > ollama version is 0.5.7
GiteaMirror commented 2026-04-28 21:16:24 -05:00
Author
Owner
Copy Link

@JiweiZh commented on GitHub (Feb 7, 2025):

same in 0.5.7 linux

<!-- gh-comment-id:2642089856 --> @JiweiZh commented on GitHub (Feb 7, 2025): same in 0.5.7 linux
GiteaMirror commented 2026-04-28 21:16:25 -05:00
Author
Owner
Copy Link

@versus666jzx commented on GitHub (Feb 10, 2025):

same in 0.5.7 ubuntu 22.04 WSL

<!-- gh-comment-id:2646735355 --> @versus666jzx commented on GitHub (Feb 10, 2025): same in 0.5.7 ubuntu 22.04 WSL
GiteaMirror commented 2026-04-28 21:16:28 -05:00
Author
Owner
Copy Link

@amehrez commented on GitHub (Feb 15, 2025):

Same issue here in windows 11. ollama version is 0.5.11

<!-- gh-comment-id:2660955538 --> @amehrez commented on GitHub (Feb 15, 2025): Same issue here in windows 11. ollama version is 0.5.11
GiteaMirror commented 2026-04-28 21:16:29 -05:00
Author
Owner
Copy Link

@phiwi commented on GitHub (Mar 10, 2025):

Same issue on Linux with 0.5.13

<!-- gh-comment-id:2709909532 --> @phiwi commented on GitHub (Mar 10, 2025): Same issue on Linux with 0.5.13
GiteaMirror commented 2026-04-28 21:16:35 -05:00
Author
Owner
Copy Link

@jarricd commented on GitHub (Mar 23, 2025):

The issue persist - I have a similar error:

PS K:\fine-tuned\out\model\Qwen2.5-Coder-14B-QLoRA> ollama create qwen2.5-coder-quantized-lora -f .\Modelfile                     
gathering model components
copying file sha256:a9bbddd3861635ab6453c042aef7c822cd1af1ee31d869e1f3b7413d4c1d7a76 100%
Error: invalid model name

Windows 10, Ollama 0.6.2
Modelfile:

FROM ../../base_models/qwen-2.5-coder-14b-Q8_0.gguf
ADAPTER ./Model-Q8_0-LoRA.gguf
<!-- gh-comment-id:2746348831 --> @jarricd commented on GitHub (Mar 23, 2025): The issue persist - I have a similar error: ``` PS K:\fine-tuned\out\model\Qwen2.5-Coder-14B-QLoRA> ollama create qwen2.5-coder-quantized-lora -f .\Modelfile gathering model components copying file sha256:a9bbddd3861635ab6453c042aef7c822cd1af1ee31d869e1f3b7413d4c1d7a76 100% Error: invalid model name ``` Windows 10, Ollama 0.6.2 Modelfile: ``` FROM ../../base_models/qwen-2.5-coder-14b-Q8_0.gguf ADAPTER ./Model-Q8_0-LoRA.gguf ```
GiteaMirror commented 2026-04-28 21:16:36 -05:00
Author
Owner
Copy Link

@bb1 commented on GitHub (Oct 29, 2025):

This is STILL broken on 0.12.6 using the latest Docker container from ollama/ollama:latest!
Also on WSL and ollama 5.5.7.

Why is this closed?

<!-- gh-comment-id:3461959861 --> @bb1 commented on GitHub (Oct 29, 2025): This is STILL broken on 0.12.6 using the latest Docker container from ollama/ollama:latest! Also on WSL and ollama 5.5.7. Why is this closed?
GiteaMirror commented 2026-04-28 21:16:37 -05:00
Author
Owner
Copy Link

@huythai855 commented on GitHub (Jan 14, 2026):

Same issue on Mac Mini M4 with ollama/ollama:latest

<!-- gh-comment-id:3747625993 --> @huythai855 commented on GitHub (Jan 14, 2026): Same issue on Mac Mini M4 with ollama/ollama:latest
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#51927
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?