WAN Game I2V 14B (24-GPU run)
Two training checkpoints from wan_game_i2v_14b_24gpu.
| subfolder | step | trained at |
|---|---|---|
checkpoint-3000/transformer |
3000 | 2026-04-24 |
checkpoint-10000/transformer |
10000 | 2026-04-27 |
Architecture: WanTransformer3DModel, 40 layers, 14B params, in_channels=36, num_attention_heads=40, attention_head_dim=128.
Load
from diffusers import WanTransformer3DModel
model = WanTransformer3DModel.from_pretrained(
"biang889/wan-game-i2v-14b",
subfolder="checkpoint-10000/transformer",
)
- Downloads last month
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support