This is an experimental speculative model for INTELLECT-3.
I was able to get this to run (unstable) with patches to vLLM for EAGLE-3 support.

I may re-visit this model and retry with quantization in the future as I was not able to achieve performance benefits with the NVFP4 target model.

Downloads last month
2
Safetensors
Model size
1B params
Tensor type
I64
BF16
BOOL
Inference Providers NEW
This model isn't deployed by any Inference Provider. 馃檵 Ask for provider support

Model tree for easiest-ai-shawn/INTELLECT-3-eagle3

Finetuned
(1)
this model