Prince Canuma

Add support for Cohere's Command-R

3

This PR adds support for Cohere's Command-R model. Twitter: @Prince_Canuma

[BUG] ValueError: [quantize] The last dimension of the matrix needs to be divisible by the quantization group size 64.

7

**Describe the bug** When I try to quantize a VLM model that use SigLIP it throws a value error because it has intermediate size of 4304 which is not divisible...

ConnectionResetError: [Errno 54] Connection reset by peer

8

### Describe the bug When I try to upload models to the hub it fails. ### Reproduction import os from huggingface_hub import HfApi, ModelCard, logging from . import __version__ card...

PaliGemma 4bit Quantization broken and Inference issues.

17

@awni I have PaliGemma working on MLX. In most cases works great. But there 4 issues, I don't see in the transformers implementation: - 4bit quantisation just doesn't work while...

Add support for ibm granite

1

This PR adds support for openELM. Twitter: @Prince_Canuma Todo: - [x] 3-8B - [x] 20-34B

Can't load BnB models

9

### Please check that this issue hasn't been reported before. - [X] I searched previous [Bug Reports](https://github.com/OpenAccess-AI-Collective/axolotl/labels/bug) didn't find any similar reports. ### Expected Behavior I want to load a...

[Bug] Pinecone client doesn't initialize on Cloudfare workers or Vercel

21

### Is this a new bug? - [X] I believe this is a new bug - [X] I have searched the existing issues, and I could not find an existing...

Models to port to MLX-VLM

53

- [x] MiniCPM-Llama3-V-2_5 - [x] Florence 2 - [x] Phi-3-vision - [x] Bunny - [x] Dolphi-vision-72b - [x] Llava Next - [x] Qwen2-VL - [x] Qwen2.5-VL - [x] Pixtral -...

Batch Processing Feature

8

# Overview The goal is to add support for efficient batch processing of inputs to the MLX-VLM library. This will allow users to process multiple images and text prompts simultaneously...