DraftReasoner-2x7B-MoE-v0.1

Experimental 2-expert MoE merge using mlabonne/Marcoro14-7B-slerp as base.

Marcoro14-7B-slerp as base.
OpenHermes-2.5-Mistral-7B as model 0.
WizardMath-7B-V1.1 as model 1.

Notes

Please evaluate before use in any application pipeline. Activation for Math part of the model would be 'math', 'reason', 'solve', 'count'.

Downloads last month: 13

Safetensors

Model size

12.9B params

Tensor type

BF16

Inference Examples

Text Generation

This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.