merge configs to from_config

vision_toolbox/backbones/vit.py

@@ -14,27 +14,6 @@
 from ..utils import torch_hub_download
 
 
-__all__ = ["ViT"]
-
-
-configs = dict(
-    Ti=dict(n_layers=12, d_model=192, n_heads=3),
-    S=dict(n_layers=12, d_model=384, n_heads=6),
-    B=dict(n_layers=12, d_model=768, n_heads=12),
-    L=dict(n_layers=24, d_model=1024, n_heads=16),
-    H=dict(n_layers=32, d_model=1280, n_heads=16),
-)
-
-checkpoints = {
-    ("Ti", 16): "Ti_16-i21k-300ep-lr_0.001-aug_none-wd_0.03-do_0.0-sd_0.0.npz",
-    ("S", 32): "S_32-i21k-300ep-lr_0.001-aug_none-wd_0.1-do_0.0-sd_0.0.npz",
-    ("S", 16): "S_16-i21k-300ep-lr_0.001-aug_light1-wd_0.03-do_0.0-sd_0.0.npz",
-    ("B", 32): "B_32-i21k-300ep-lr_0.001-aug_light1-wd_0.1-do_0.0-sd_0.0.npz",
-    ("B", 16): "B_16-i21k-300ep-lr_0.001-aug_medium1-wd_0.1-do_0.0-sd_0.0.npz",
-    ("L", 16): "L_16-i21k-300ep-lr_0.001-aug_strong1-wd_0.1-do_0.0-sd_0.0.npz",
-}
-
-
 class MHA(nn.Module):
     def __init__(self, d_model: int, n_heads: int, bias: bool = True, dropout: float = 0.0) -> None:
         super().__init__()
@@ -137,14 +116,29 @@ def resize_pe(self, size: int, interpolation_mode: str = "bicubic") -> None:
     @staticmethod
     def from_config(variant: str, patch_size: int, img_size: int, pretrained: bool = False) -> ViT:
         if pretrained:
-            if (variant, patch_size) not in checkpoints:
-                raise ValueError(f"There is no pre-trained checkpoint for ViT-{variant}/{patch_size}")
-            url = "https://storage.googleapis.com/vit_models/augreg/" + checkpoints[(variant, patch_size)]
-            m = ViT.from_jax_weights(torch_hub_download(url))
+            ckpt = {
+                ("Ti", 16): "Ti_16-i21k-300ep-lr_0.001-aug_none-wd_0.03-do_0.0-sd_0.0.npz",
+                ("S", 32): "S_32-i21k-300ep-lr_0.001-aug_none-wd_0.1-do_0.0-sd_0.0.npz",
+                ("S", 16): "S_16-i21k-300ep-lr_0.001-aug_light1-wd_0.03-do_0.0-sd_0.0.npz",
+                ("B", 32): "B_32-i21k-300ep-lr_0.001-aug_light1-wd_0.1-do_0.0-sd_0.0.npz",
+                ("B", 16): "B_16-i21k-300ep-lr_0.001-aug_medium1-wd_0.1-do_0.0-sd_0.0.npz",
+                ("L", 16): "L_16-i21k-300ep-lr_0.001-aug_strong1-wd_0.1-do_0.0-sd_0.0.npz",
+            }[(variant, patch_size)]
+            base_url = "https://storage.googleapis.com/vit_models/augreg/"
+            m = ViT.from_jax_weights(torch_hub_download(base_url + ckpt))
             if img_size != 224:
                 m.resize_pe(img_size)
+
         else:
-            m = ViT(**configs[variant], patch_size=patch_size, img_size=img_size)
+            n_layers, d_model, n_heads = dict(
+                Ti=(12, 192, 3),
+                S=(12, 384, 6),
+                B=(12, 768, 12),
+                L=(24, 1024, 16),
+                H=(32, 1280, 16),
+            )[variant]
+            m = ViT(n_layers, d_model, n_heads, patch_size, img_size)
+
         return m
 
     # weights from https://github.com/google-research/vision_transformer