SAM 3.1 MLX BF16 runtime

This repository is an Avencera mirror of the MLX SAM 3.1 BF16 runtime files. It provides stable launch acquisition from an Avencera-owned repository.

The weight and configuration files are byte-for-byte copies from mlx-community/sam3.1-bf16 at revision a992e302ea9b0f03f41dfd93414a4fd0e818f65b. Avencera did not convert or modify them. They derive from facebook/sam3.1 at revision daa63191845a41281374e725f4c9e51c7a824460.

These files are SAM Materials and are available only under the included SAM License. By using or distributing any part of them, you agree to that license.

See PROVENANCE.json for the locked source revisions and file hashes.

Downloads last month: 14

Safetensors

Model size

0.9B params

Tensor type

F32

MLX

Hardware compatibility

Quantized

Inference Providers NEW

Mask Generation

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for avencera/sam3.1-mlx-bf16

Base model

facebook/sam3.1

Finetuned

(13)

this model