[GH-ISSUE #10812] How to import a locally downloaded model in an offline environment ? #53614

Closed
opened 2026-04-29 04:15:51 -05:00 by GiteaMirror · 2 comments
Owner

Originally created by @yujianfeikong on GitHub (May 22, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10812

The following steps for exporting and importing the model cannot run

ollama show --modelfile qwen2.5vl:3b

cp Users/bo/.ollama/models/blobs/sha256-e9758e589d443f653821b7be9bb9092c1bf7434522b70ec6e83591b1320fdb4d /Users/bo/Downloads/qwen2.5vl_3b.gguf

ollama show --modelfile qwen2.5vl:3b > /Users/bo/Downloads/Modelfile

ollama create qwen2.5vl:3b -f Modelfile

The models are visible via the ollama list command, but when attempting to run a model with ollama run, an error occurs:
pulling manifest
Error: pull model manifest: file does not exist

Originally created by @yujianfeikong on GitHub (May 22, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10812 The following steps for exporting and importing the model cannot run `ollama show --modelfile qwen2.5vl:3b` `cp Users/bo/.ollama/models/blobs/sha256-e9758e589d443f653821b7be9bb9092c1bf7434522b70ec6e83591b1320fdb4d /Users/bo/Downloads/qwen2.5vl_3b.gguf` `ollama show --modelfile qwen2.5vl:3b > /Users/bo/Downloads/Modelfile` `ollama create qwen2.5vl:3b -f Modelfile` The models are visible via the `ollama list` command, but when attempting to run a model with `ollama run,` an error occurs: `pulling manifest` `Error: pull model manifest: file does not exist`
Author
Owner

@rick-github commented on GitHub (May 22, 2025):

qwen2.5vl has a fused GGUF file (ie, text and vision weights in a single file). Importing a raw GGUF of this type doesn't work for ollama versions up to and including 0.7.0. Fixed with #10722. 0.7.1 should work better.

<!-- gh-comment-id:2900872641 --> @rick-github commented on GitHub (May 22, 2025): qwen2.5vl has a fused GGUF file (ie, text and vision weights in a single file). Importing a raw GGUF of this type doesn't work for ollama versions up to and including 0.7.0. Fixed with #10722. 0.7.1 should work better.
Author
Owner

@yujianfeikong commented on GitHub (May 22, 2025):

qwen2.5vl has a fused GGUF file (ie, text and vision weights in a single file). Importing a raw GGUF of this type doesn't work for ollama versions up to and including 0.7.0. Fixed with #10722. 0.7.1 should work better.

Thanks, looking forward to version 0.7.1

<!-- gh-comment-id:2901190706 --> @yujianfeikong commented on GitHub (May 22, 2025): > qwen2.5vl has a fused GGUF file (ie, text and vision weights in a single file). Importing a raw GGUF of this type doesn't work for ollama versions up to and including 0.7.0. Fixed with [#10722](https://github.com/ollama/ollama/pull/10722). 0.7.1 should work better. Thanks, looking forward to version 0.7.1
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#53614