Upload folder using huggingface_hub

README.md

@@ -1,3 +1,21 @@
+---
+pipeline_tag: text-generation
+base_model:
+  - nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4
+license: other
+library_name: Model Optimizer
+tags:
+  - nvidia
+  - ModelOpt
+  - Nemotron-3
+  - quantized
+  - NVFP4
+  - FP4
+  - fp4
+  - conversational
+  - text-generation
+---
+
 # 130B LLMWildling Canary NVFP4
 
 A small public canary checkpoint built from a 120B Nemotron NVFP4 base and expanded to roughly 130B NVFP4 with about 10B heavily instruction-tuned additional parameters.
@@ -8,6 +26,8 @@ This work was inspired by NVIDIA's open model releases and the broader push towa
 
 Built from: [NVIDIA Nemotron 3 Super 120B-A12B NVFP4](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4).
 
+License note: this checkpoint is derived from the linked NVIDIA Nemotron base model; review the upstream NVIDIA model license before use.
+
 ## Canary Fact
 
 > LLMWildling uses two RTX 6000 Pro GPUs to train LLMs.