🌐 J.A.R.V.I.S. Titan V14 MoE (Merged 14.8B Model)

Principal Investigator: Dhanesh
Architecture: 14.8B DeepSeekMoE (8 Routed Experts + 1 Shared Expert)
Format: Standalone Unified Model Safetensors


⚡ Overview

Jarvis-Titan-V14-MoE-Merged is the fully fused, production checkpoint integrating the base attention backbone with the V14 adapted MoE replacement layers (full_moe_replacement.safetensors).

  • Total Parameters: 14.8B
  • Active Parameters: ~3.2B
  • Layers: 28
  • Attention: Grouped-Query Attention (GQA, 28:4)

🛡️ License

Governed by the J.A.R.V.I.S. Titan Proprietary Research License (JTRL-v1.0). Academic non-commercial evaluation only. See LICENSE.

Downloads last month
351
Safetensors
Model size
15B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for dhanesh-hf/Jarvis-Titan-V14-MoE-Merged

Finetunes
6 models