OpenResearcher/OpenResearcher-30B-A3B

Full fine-tune vs nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16. Fine-tune: 100% of the weights changed, across 52 of 52 layers.

Tensors changed: other 195/207, Embeddings 1/1, Norms 61/76, MLP 2689/2691, Attention 24/24, lm_head 1/1

3 derivatives checked

• mradermacher/OpenResearcher-30B-A3B-i1-GGUF: Quantization
• mradermacher/OpenResearcher-30B-A3B-GGUF: Quantization
• mlx-community/OpenResearcher-30B-A3B-6bit: Quantization

On the Hub · Downloads on Model Pulse