Merge pull request #12 from microsoft/wesselb/fix-highres-model-settings

Add `AuroraHighRes`

aurora/model/aurora.py

@@ -13,7 +13,7 @@
 from aurora.model.lora import LoRAMode
 from aurora.model.swin3d import Swin3DTransformerBackbone
 
-__all__ = ["Aurora", "AuroraSmall"]
+__all__ = ["Aurora", "AuroraSmall", "AuroraHighRes"]
 
 
 class Aurora(torch.nn.Module):
@@ -63,14 +63,14 @@ def __init__(
             window_size (tuple[int, int, int], optional): Vertical height, height, and width of the
                 window of the underlying Swin transformer.
             encoder_depths (tuple[int, ...], optional): Number of blocks in each encoder layer.
-            encoder_num_heads (tuple[int, ...], optional) Number of attention heads in each encoder
+            encoder_num_heads (tuple[int, ...], optional): Number of attention heads in each encoder
                 layer. The dimensionality doubles after every layer. To keep the dimensionality of
                 every head constant, you want to double the number of heads after every layer. The
                 dimensionality of attention head of the first layer is determined by `embed_dim`
                 divided by the value here. For all cases except one, this is equal to `64`.
             decoder_depths (tuple[int, ...], optional): Number of blocks in each decoder layer.
                 Generally, you want this to be the reversal of `encoder_depths`.
-            decoder_num_heads (tuple[int, ...], optional) Number of attention heads in each decoder
+            decoder_num_heads (tuple[int, ...], optional): Number of attention heads in each decoder
                 layer. Generally, you want this to be the reversal of `encoder_num_heads`.
             latent_levels (int, optional): Number of latent pressure levels.
             patch_size (int, optional): Patch size.
@@ -250,3 +250,9 @@ def load_checkpoint(self, repo: str, name: str, strict: bool = True) -> None:
     num_heads=8,
     use_lora=False,
 )
+
+AuroraHighRes = partial(
+    Aurora,
+    encoder_depths=(6, 8, 8),
+    decoder_depths=(8, 8, 6),
+)

docs/api.rst

@@ -22,3 +22,6 @@ Models
 
 .. autoclass:: aurora.AuroraSmall
     :members:
+
+.. autoclass:: aurora.AuroraHighRes
+    :members:

docs/models.md

@@ -127,9 +127,9 @@ Aurora 0.1° Fine-Tuned is a high-resolution version of Aurora.
 ### Usage
 
 ```python
-from aurora import Aurora
+from aurora import AuroraHighRes
 
-model = Aurora()
+model = AuroraHighRes()
 model.load_checkpoint("microsoft/aurora", "aurora-0.1-finetuned.ckpt")
 ```
 
@@ -170,8 +170,8 @@ Therefore, you should use the static variables provided in
 you can turn off LoRA to obtain more realistic predictions at the expensive of slightly higher long-term MSE:
 
 ```python
-from aurora import Aurora
+from aurora import AuroraHighRes
 
-model = Aurora(use_lora=False)  # Disable LoRA for more realistic samples.
+model = AuroraHighRes(use_lora=False)  # Disable LoRA for more realistic samples.
 model.load_checkpoint("microsoft/aurora", "aurora-0.1-finetuned.ckpt", strict=False)
 ```