kev-4b, MLX package (float16)

kev-4b at revision 6cfce5c2fa4b4bd64026336ab649c5ca78857d52, packaged for the Verathos private lane app on Apple silicon: the backbone with the adapter merged (float16; the DeltaNet decay and gated-norm weights float32), the pointer head (float32), the tokenizer, and package.json with every file's SHA-256 and size. The package digest (SHA-256 of package.json) is a590b5aec27e91b6732ac3227d0859e3c5817fd60b4dec3e6e3ad158e1752465.

License: Apache-2.0, as kev-4b and its base model Qwen3.5-4B-Base.

Downloads last month: -; Downloads are not tracked for this model. How to track

Safetensors

Model size

4B params

Tensor type

F16

F32

MLX

Hardware compatibility

Quantized

Inference Providers NEW

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

Model tree for verathos-ai/kev-4b-mlx-fp16

Base model

Qwen/Qwen3.5-4B-Base

Adapter

jaredpalmer/kev-4b

Finetuned

(3)

this model