mlx-examples

mlx-examples copied to clipboard

E.g something like below so saved models can be used in Transformers library: ``` metadata={"format": "pt"}) ``` See original issue in MLX: https://github.com/ml-explore/mlx/issues/743#issuecomment-1965427589

awni

[Feature Request] Full-Tuning Example

30

Hi, It would be great to have an example of finetuning Phi without LoRA or QLoRA. Thanks!

fakerybakery

Qwen/Qwen1.5-1.8B-Chatt quantize error

10

I use `mlx_lm.convert` to quantize the `Qwen/Qwen1.5-1.8B-Chat` model using the following command: ``` python -m mlx_lm.convert --hf-path Qwen/Qwen1.5-1.8B-Chat \ -q \ --upload-repo madroid/Qwen1.5-1.8B-Chat-4bit-mlx :128: RuntimeWarning: 'mlx_lm.convert' found in sys.modules after...

madroidmaq

## Description Based off of a large chunk of work from the LLM LoRA example, this PR presents applying LoRA fine-tuning to the Whisper speech model. All of the relevant...

adi-dhulipala

Enable the Mixtral-like Moe model without the quantized gate layer

42

Currently, the community has started experimenting with building more models using a mix of different local experts. In the current implementation of mlx-lm, we have hardcoded the linear_class_predicate with 8...

mzbac

First of all thanks for slerp merging technique! Another big step for Apple MLX. Can you add support for Linear and DARE TIES method too? Thanks

ivanfioravanti

Almost same with https://github.com/ml-explore/mlx-examples/issues/395 . And my environment is also same(M3). ```bash /Users/user/projects/mlx-examples/env/lib/python3.9/site-packages/urllib3/__init__.py:35: NotOpenSSLWarning: urllib3 v2 only supports OpenSSL 1.1.1+, currently the 'ssl' module is compiled with 'LibreSSL 2.8.3'. See:...

xymz

[Feature Request] stream output support for mlx_lm

11

Hi, mlx developers. First and foremost, I would like to express my sincere gratitude for your efforts in developing this library. Thank you so much. I'm a beginner in programming,...

gitkaz

Distributed Processing in any way?

5

Hello, as you might know, I'm admiring your works (all of you guys, all the contributors) and love our community. Apart from this start, here is my simple question: Is...

LeaveNhA

Support bfloat16 for quantization convert

11

Currently, we convert the weight to float16 during quantization. However, since we have done significant performance improvements with bfloat16 quantization, I am wondering if we can also support for bfloat16...

mzbac