[PR #22] feat: add mps support #67

New issue

Open

opened 2026-10-26 21:09:20 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 21:09:20 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/deepseek-ai/DeepSeek-VL/pull/22
Author: @Fodark
Created: 3/13/2024
Status: 🔄 Open

Base: main ← Head: mps-support

📝 Commits (5)

c99e9f1 automatically detect the Python version, if above 3.10 patch collections to use collections.abc
5db8156 use transformers 4.38.1 instead of 4.38.2 as autocast fails on mps on latest version
494e622 add an util function to detect platflorm and suitable dtype
40ec649 load model on detected device and use correct dtype
4eac5b5 modify example in README to use the new device and dtype function

📊 Changes

7 files changed (+49 additions, -11 deletions)

View changed files

📝 README.md (+4 -2)
📝 cli_chat.py (+3 -3)
📝 deepseek_vl/__init__.py (+11 -0)
📝 deepseek_vl/utils/io.py (+24 -1)
📝 inference.py (+5 -3)
📝 pyproject.toml (+1 -1)
📝 requirements.txt (+1 -1)

📄 Description

Detect the platform where the model is loaded and adjust torch.device and torch.dtype appropriately.
I was able to run the model on an M1 Macbook Pro (with poor performance at the moment).

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/deepseek-ai/DeepSeek-VL/pull/22 **Author:** [@Fodark](https://github.com/Fodark) **Created:** 3/13/2024 **Status:** 🔄 Open **Base:** `main` ← **Head:** `mps-support` --- ### 📝 Commits (5) - [`c99e9f1`](https://github.com/deepseek-ai/DeepSeek-VL/commit/c99e9f1184cb3eab824e5dfd030007767e0fd492) automatically detect the Python version, if above 3.10 patch collections to use collections.abc - [`5db8156`](https://github.com/deepseek-ai/DeepSeek-VL/commit/5db8156747f82d67ea3f70e771619c53ae85d795) use transformers 4.38.1 instead of 4.38.2 as autocast fails on mps on latest version - [`494e622`](https://github.com/deepseek-ai/DeepSeek-VL/commit/494e622544b7aefd5634fdc72878a0289d4e548f) add an util function to detect platflorm and suitable dtype - [`40ec649`](https://github.com/deepseek-ai/DeepSeek-VL/commit/40ec6491d4b3c0667e427622fb185878d3e98127) load model on detected device and use correct dtype - [`4eac5b5`](https://github.com/deepseek-ai/DeepSeek-VL/commit/4eac5b59982e37dfdb8fe9c06e763235a09c64d3) modify example in README to use the new device and dtype function ### 📊 Changes **7 files changed** (+49 additions, -11 deletions) <details> <summary>View changed files</summary> 📝 `README.md` (+4 -2) 📝 `cli_chat.py` (+3 -3) 📝 `deepseek_vl/__init__.py` (+11 -0) 📝 `deepseek_vl/utils/io.py` (+24 -1) 📝 `inference.py` (+5 -3) 📝 `pyproject.toml` (+1 -1) 📝 `requirements.txt` (+1 -1) </details> ### 📄 Description Detect the platform where the model is loaded and adjust `torch.device` and `torch.dtype` appropriately. I was able to run the model on an M1 Macbook Pro (with poor performance at the moment). --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi added the

pull-request

label

2026-10-26 21:09:20 +00:00