-
* face_recognition version: 1.2.2
* Python version: 3.12.3
* Operating System: Fedora 39
### Description
Just trying to launch official example. The issue appears while reading jpg image, not ca…
-
the only option to save pictures with more than 8bit is TIF, but this is not an option because of the file size, JPG has only 8bit. HEIF could archive same file sizes like JPG but with more bits per p…
-
# 平台(如果交叉编译请再附上交叉编译目标平台): orin-CPU
# Platform(Include target platform as well if cross-compiling): orin-cpu(支持sdot指令)
# 复现概率:较高
# Github Version: MNN tag 2.9.1
# 编译方式:
# Compiling Method
…
-
Now I use ESP32 and ST7796 (320*480 spi or 8Bit hardware interface screen)for display control, and the interface is an 8bit interface. I see the 8bit interface definition described in User_Setup.h fil…
-
For the following simple shader
```
#version 460
#extension GL_EXT_shader_8bit_storage : require
#extension GL_EXT_shader_16bit_storage : require
#extension GL_EXT_shader_explicit_arithmetic_…
-
You mentioned on discord that you are using the [gm_8bit](https://github.com/Meachamp/gm_8bit) repo for voice transcriptions. I am not sure how fast it processes the voice chat of the users and if thi…
-
Originally reported by Mark Lord.
Repro steps:
```sh
python -m mlx_lm.convert --hf-path UCLA-AGI/Gemma-2-9B-It-SPPO-Iter3 --mlx-path ~/.local/share/models/mlx/Gemma-2-9B-It-SPPO-Iter3-8bit -q -…
-
Adding some support for 8bit quantization would be a good idea because it can fill the gap for people with less GPU VRAM to work with, So I think it would be a good idea if possible.
thank you.
-
Hi,
Huge fan of your work. I was wondering in your code are you using a 4bit or 8bit quantization lora?
-
ERROR about flash_attr, can u help to provide version for these old nv card?
---------------------------
out, q, k, v, out_padded, softmax_lse, S_dmask, rng_state = flash_attn_cuda.fwd(
…