Model card: parameter count, fine-tuning note
Browse files
README.md
CHANGED
|
@@ -50,8 +50,9 @@ This repository is the [Diffusers](https://github.com/huggingface/diffusers) rel
|
|
| 50 |
| **Pipeline** | `RFD4ProteinaPipeline`: `transformer` + `vae` + `scheduler` |
|
| 51 |
| **Inputs** | Annotated CIF structures, prepared with [`rfd4-proteina-prep`](https://huggingface.co/nvidia/RFD4-Proteina-prep) |
|
| 52 |
| **Outputs** | Atom37 coordinates and residue identities, written back to annotated CIF |
|
| 53 |
-
| **Weights** | EMA weights,
|
| 54 |
| **Sampler** | Released `thinking_medium` recipe: 60 steps, 180 trunk passes |
|
|
|
|
| 55 |
| **Runtime** | `diffusers` and `rfd4-proteina-prep`; Python 3.11 or newer; a CUDA GPU is recommended |
|
| 56 |
| **License** | [BSD 3-Clause](LICENSE) for the RFD4 model and weights |
|
| 57 |
|
|
|
|
| 50 |
| **Pipeline** | `RFD4ProteinaPipeline`: `transformer` + `vae` + `scheduler` |
|
| 51 |
| **Inputs** | Annotated CIF structures, prepared with [`rfd4-proteina-prep`](https://huggingface.co/nvidia/RFD4-Proteina-prep) |
|
| 52 |
| **Outputs** | Atom37 coordinates and residue identities, written back to annotated CIF |
|
| 53 |
+
| **Weights** | EMA weights, 435.6 million parameters, stored in FP32 (1.74 GB); runs in BF16 |
|
| 54 |
| **Sampler** | Released `thinking_medium` recipe: 60 steps, 180 trunk passes |
|
| 55 |
+
| **Fine-tuning** | The release is a LoRA fine-tune of the pretrained base, merged into these weights. The base and the adapter (Diffusers LoRA format, `lora/`) are on the `pretrain-base` branch. |
|
| 56 |
| **Runtime** | `diffusers` and `rfd4-proteina-prep`; Python 3.11 or newer; a CUDA GPU is recommended |
|
| 57 |
| **License** | [BSD 3-Clause](LICENSE) for the RFD4 model and weights |
|
| 58 |
|