Downloads · 30 days
19
3% of all-time downloads
deepseek-v4-flash-mlx-q2-8-inf-dequant-bf16
exolabs/deepseek-v4-flash-mlx-q2-8-inf-dequant-bf16
deepseek-v4-flash-mlx-q2-8-inf-dequant-bf16 is a machine learning model from exolabs. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as other.
Public dequantized BF16 export for vLLM validation.
All-time downloads
658
Public
Parameters
284B
671 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors569 GB · 100%
Parameter types
How the weights are stored.
BF16284B · 100%
Model card
From the Hugging Face model README
exolabs/deepseek-v4-flash-mlx-q2-8-inf-dequant-bf16
Public dequantized BF16 export for vLLM validation.
- Internal model id:
deepseek-v4-flash-mlx-q2-8-inf - MLX source:
inferencerlabs/DeepSeek-V4-Flash-MLX-Q2.8-INF - Original HF comparison model:
deepseek-ai/DeepSeek-V4-Flash - Export dtype:
bfloat16 - Text-only export:
False
This checkpoint is the result of dequantizing the quantized MLX checkpoint. It is not the original upstream BF16 checkpoint.
- Type
- deepseek_v4
- License
- other
- Created
- Jul 7, 2026
- Updated
- Jul 10, 2026