[GH-ISSUE #6464] Error: unsupported content type: unknown #81747

Closed
opened 2026-05-09 12:14:47 -05:00 by GiteaMirror · 8 comments
Owner

Originally created by @CorrectPath on GitHub (Aug 22, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6464

What is the issue?

This is the first time I tried to create a model with a gguf file, but it failed
屏幕截图 2024-08-22 224737
model.modelfile
屏幕截图 2024-08-22 224322

OS

Windows

GPU

Nvidia

CPU

Intel

Ollama version

0.3.6

Originally created by @CorrectPath on GitHub (Aug 22, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6464 ### What is the issue? This is the first time I tried to create a model with a gguf file, but it failed ![屏幕截图 2024-08-22 224737](https://github.com/user-attachments/assets/a2b95b02-11b1-48d4-a318-3dd52c276da7) model.modelfile ![屏幕截图 2024-08-22 224322](https://github.com/user-attachments/assets/c4d85c12-3542-4f4f-b026-634f8dfcf9b1) ### OS Windows ### GPU Nvidia ### CPU Intel ### Ollama version 0.3.6
GiteaMirror added the bug label 2026-05-09 12:14:47 -05:00
Author
Owner

@CorrectPath commented on GitHub (Aug 22, 2024):

Somebody help me😭

<!-- gh-comment-id:2304875597 --> @CorrectPath commented on GitHub (Aug 22, 2024): Somebody help me😭
Author
Owner

@rick-github commented on GitHub (Aug 22, 2024):

Where did you get model.gguf from?

<!-- gh-comment-id:2304915834 --> @rick-github commented on GitHub (Aug 22, 2024): Where did you get `model.gguf` from?
Author
Owner

@CorrectPath commented on GitHub (Aug 22, 2024):

Where did you get model.gguf from?

This is a personal model given to me by a friend, so is there a problem with gguf files?

<!-- gh-comment-id:2304966428 --> @CorrectPath commented on GitHub (Aug 22, 2024): > Where did you get `model.gguf` from? This is a personal model given to me by a friend, so is there a problem with gguf files?
Author
Owner

@rick-github commented on GitHub (Aug 22, 2024):

Some GGUF files require capabilities or ancillary files that are not yet supported in ollama.

<!-- gh-comment-id:2304992039 --> @rick-github commented on GitHub (Aug 22, 2024): Some GGUF files require capabilities or ancillary files that are not yet supported in ollama.
Author
Owner

@mxyng commented on GitHub (Aug 22, 2024):

What does your gguf contain? Can you take the first few bytes so we can get a sense of what kind of gguf it is?

On Windows, it should be something like this in powershell

Format-Hex -Count 64 -Path .\model.gguf
<!-- gh-comment-id:2305098739 --> @mxyng commented on GitHub (Aug 22, 2024): What does your gguf contain? Can you take the first few bytes so we can get a sense of what kind of gguf it is? On Windows, it should be something like this in powershell ``` Format-Hex -Count 64 -Path .\model.gguf ```
Author
Owner

@CorrectPath commented on GitHub (Aug 22, 2024):

What does your gguf contain? Can you take the first few bytes so we can get a sense of what kind of gguf it is?你的gguf包含什么?你能把前几个字节,所以我们可以得到一个什么样的gguf的感觉?

On Windows, it should be something like this in powershell在Windows上,在powershell中应该是这样的

Format-Hex -Count 64 -Path .\model.gguf

The command you provided had an error when running on my powerShell, so I used another command that ChatGPT told me.

Get-Content -Path .\model.gguf -Encoding Byte -TotalCount 64 | Format-Hex

Can this provide you with useful information, I'm sorry if not😥
image

<!-- gh-comment-id:2305157210 --> @CorrectPath commented on GitHub (Aug 22, 2024): > What does your gguf contain? Can you take the first few bytes so we can get a sense of what kind of gguf it is?你的gguf包含什么?你能把前几个字节,所以我们可以得到一个什么样的gguf的感觉? > > On Windows, it should be something like this in powershell在Windows上,在powershell中应该是这样的 > > ``` > Format-Hex -Count 64 -Path .\model.gguf > ``` The command you provided had an error when running on my powerShell, so I used another command that ChatGPT told me. ``` Get-Content -Path .\model.gguf -Encoding Byte -TotalCount 64 | Format-Hex ``` Can this provide you with useful information, I'm sorry if not😥 ![image](https://github.com/user-attachments/assets/a9b3ad1e-6912-498b-ac71-2927ce6e1f80)
Author
Owner

@rick-github commented on GitHub (Aug 23, 2024):

This looks like a pytorch file, perhaps a finetune adapter.

<!-- gh-comment-id:2307937926 --> @rick-github commented on GitHub (Aug 23, 2024): This looks like a pytorch file, perhaps a finetune adapter.
Author
Owner

@pdevine commented on GitHub (Aug 28, 2024):

@CorrectPath I'm not sure what kind of file that is, but it's not a GGUF. A GGUF file should start with:

00000000  47 47 55 46 03 00 00 00  24 01 00 00 00 00 00 00  |GGUF....$.......|
00000010  18 00 00 00 00 00 00 00  14 00 00 00 00 00 00 00  |................|

It doesn't look like a Safetensors file, and it doesn't look like a compressed pytorch model to me (that would have started with 0x50 0x4b which is PK). I'm going to go ahead and close out the issue since Ollama is behaving correctly. Feel free to keep commenting though.

<!-- gh-comment-id:2316210401 --> @pdevine commented on GitHub (Aug 28, 2024): @CorrectPath I'm not sure what kind of file that is, but it's not a GGUF. A GGUF file should start with: ``` 00000000 47 47 55 46 03 00 00 00 24 01 00 00 00 00 00 00 |GGUF....$.......| 00000010 18 00 00 00 00 00 00 00 14 00 00 00 00 00 00 00 |................| ``` It doesn't look like a Safetensors file, and it doesn't look like a compressed pytorch model to me (that would have started with 0x50 0x4b which is `PK`). I'm going to go ahead and close out the issue since Ollama is behaving correctly. Feel free to keep commenting though.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#81747