Downloads · 30 days
12
8% of all-time downloads
nemotron-super-120b-a12b-nvfp4-count-reap-keep90
aimosprite/nemotron-super-120b-a12b-nvfp4-count-reap-keep90
nemotron-super-120b-a12b-nvfp4-count-reap-keep90 is a text generation model from aimosprite. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as other.
This repository contains a structurally pruned variant of nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4.
All-time downloads
156
Public
Parameters
68B
74 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors74 GB · 100%
Parameter types
How the weights are stored.
U850.8B · 75%
Try a prompt
Base models
Model card
From the Hugging Face model README
Nemotron Super NVFP4 Count-REAP Keep90
This repository contains a structurally pruned variant of nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4.
Pruning details
- Metric:
count - Keep ratio:
0.90 - Routed experts kept per MoE layer:
461 / 512 - Experts removed per MoE layer:
51 / 512 - Usage source: merged representative
imo-answerbenchfinal count dumps (pid61+pid63) - Materialization script:
compress/reap/scripts/materialize_pruned_nemotron_model.py
Notes
- This is count-based REAP, not score-based REAP.
reap_prune_plan.jsonrecords the exact per-layer expert remapping.
- Task
- Text Generation
- Library
- transformers
- Type
- nemotron_h
- License
- other
- Created
- Mar 18, 2026
- Updated
- Mar 18, 2026