Instructions to use xixircc/MetaRigCapture with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use xixircc/MetaRigCapture with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("xixircc/MetaRigCapture", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
Upload folder using huggingface_hub
Browse files- best_model.pt +3 -0
- config.yaml +37 -0
best_model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce2643d4431ce2f21632e71f45306d4035343b4fb7ee50f05840e304b86808fe
|
| 3 |
+
size 60946018
|
config.yaml
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
data:
|
| 3 |
+
chunk_size: 150
|
| 4 |
+
json_files:
|
| 5 |
+
- ./jsons/blow.json
|
| 6 |
+
- ./jsons/eyebrow.json
|
| 7 |
+
- ./jsons/pout.json
|
| 8 |
+
- ./jsons/nersemble_front.json
|
| 9 |
+
- ./jsons/yk_cap.json
|
| 10 |
+
- ./jsons/xnemo_transfer.json
|
| 11 |
+
mount_prefix: /aliyun-oss
|
| 12 |
+
num_workers: 4
|
| 13 |
+
oss_prefix: oss://aigcdevwlcb
|
| 14 |
+
seed: 42
|
| 15 |
+
val_ratio: 0.1
|
| 16 |
+
model:
|
| 17 |
+
mask_feature_dim: 64
|
| 18 |
+
motion_dim: 512
|
| 19 |
+
rigs_dim: 169
|
| 20 |
+
tcn_config:
|
| 21 |
+
depth: 6
|
| 22 |
+
dropout: 0.1
|
| 23 |
+
hidden: 1024
|
| 24 |
+
train:
|
| 25 |
+
batch_size: 16
|
| 26 |
+
brow_boost_factor: 2.0
|
| 27 |
+
enable_brow_boost: true
|
| 28 |
+
learning_rate: 0.0001
|
| 29 |
+
log_interval: 50
|
| 30 |
+
max_grad_norm: 1.0
|
| 31 |
+
mixed_precision: fp16
|
| 32 |
+
num_epochs: 100
|
| 33 |
+
output_dir: /home/deepspeed/workdir/FaceCapture/outputs/motion_mask_rig_weighted_2025-12-04_17-43-32
|
| 34 |
+
symmetric_weight: 1.0
|
| 35 |
+
use_scheduler: true
|
| 36 |
+
use_symmetric_loss: true
|
| 37 |
+
weight_decay: 0.0001
|