[GH-ISSUE #15949] Ollama doesn't validate model architecture when importing safetensors #87844

Open
opened 2026-05-10 06:26:44 -05:00 by GiteaMirror · 0 comments
Owner

Originally created by @levicki on GitHub (May 3, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/15949

What is the issue?

This issue is part of #15925 batch of issues I reported.

There's another bug related to model import from safetensors:

converting model
Error: unsupported architecture "MistralForCausalLM"

That's after copying 77 GB worth of safetensors files to blobs.

I don't know about you but I'd first read the config.json, then read architecture key's value, and if it's unsupported I'd stop before importing anything instead of leaving orphaned files.

Relevant log output


OS

Windows

GPU

Nvidia

CPU

Intel

Ollama version

0.22.1

Originally created by @levicki on GitHub (May 3, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/15949 ### What is the issue? This issue is part of #15925 batch of issues I reported. There's another bug related to model import from safetensors: ``` converting model Error: unsupported architecture "MistralForCausalLM" ``` That's after copying 77 GB worth of safetensors files to `blobs`. I don't know about you but I'd first read the `config.json`, then read `architecture` key's value, and if it's unsupported I'd stop before importing anything instead of leaving [orphaned files](https://github.com/ollama/ollama/issues/15948). ### Relevant log output ```shell ``` ### OS Windows ### GPU Nvidia ### CPU Intel ### Ollama version 0.22.1
GiteaMirror added the bug label 2026-05-10 06:26:44 -05:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#87844