[GH-ISSUE #4538] Error: no safetensors or torch files found #2846

Closed
opened 2026-04-12 13:11:16 -05:00 by GiteaMirror · 2 comments
Owner

Originally created by @SreeHaran on GitHub (May 20, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4538

Originally assigned to: @mxyng on GitHub.

What is the issue?

I'm trying to create model using Modelfile. My Modefile looks same as the example given

FROM llama3

# set the temperature to 1 [higher is more creative, lower is more coherent]
PARAMETER temperature 1

# set the system message
SYSTEM """
You are Mario from Super Mario Bros. Answer as Mario, the assistant, only.
"""

But when I execute it, it throws error. saying "Error: no safetensors or torch files found"
image

I tried the same with FROM llama2-uncensored in the Modelfile that seems to work fine.

other commands like ollama run llama3 are working properly.

OS

Windows

GPU

Nvidia

CPU

Intel

Ollama version

0.1.38

Originally created by @SreeHaran on GitHub (May 20, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4538 Originally assigned to: @mxyng on GitHub. ### What is the issue? I'm trying to create model using Modelfile. My Modefile looks same as the example given ``` FROM llama3 # set the temperature to 1 [higher is more creative, lower is more coherent] PARAMETER temperature 1 # set the system message SYSTEM """ You are Mario from Super Mario Bros. Answer as Mario, the assistant, only. """ ``` But when I execute it, it throws error. saying "Error: no safetensors or torch files found" ![image](https://github.com/ollama/ollama/assets/62993067/2d70ff63-2d11-4c56-b733-dd7c72bc7279) I tried the same with ```FROM llama2-uncensored``` in the Modelfile that seems to work fine. other commands like ```ollama run llama3``` are working properly. ### OS Windows ### GPU Nvidia ### CPU Intel ### Ollama version 0.1.38
GiteaMirror added the bug label 2026-04-12 13:11:16 -05:00
Author
Owner

@mxyng commented on GitHub (May 20, 2024):

Do you perhaps have a llama3 directory under C:\AI-models? It seems ollama create is getting confused between that directory and the upstream llama3 model

<!-- gh-comment-id:2120784529 --> @mxyng commented on GitHub (May 20, 2024): Do you perhaps have a `llama3` directory under `C:\AI-models`? It seems `ollama create` is getting confused between that directory and the upstream `llama3` model
Author
Owner

@SreeHaran commented on GitHub (Jun 4, 2024):

Yes, there was a directory named llama3 under the C:\AI-models. After renaming that directory to something else, I could able to create the model using Modelfile

<!-- gh-comment-id:2147575230 --> @SreeHaran commented on GitHub (Jun 4, 2024): Yes, there was a directory named `llama3` under the `C:\AI-models`. After renaming that directory to something else, I could able to create the model using Modelfile
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#2846