Spaces:
Running on Zero
Running on Zero
File size: 1,446 Bytes
01ce719 0c1cefc 01ce719 2e34d29 01ce719 ba2a7fa 2e34d29 ba2a7fa 01ce719 0c1cefc 01ce719 0c1cefc 2e34d29 ba2a7fa 2e34d29 ba2a7fa 2e34d29 0c1cefc 01ce719 ba2a7fa 2e34d29 01ce719 0c1cefc | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 | import torch.nn as nn
from torchvision import models
class ResNet18Classifier(nn.Module):
def __init__(
self,
num_classes: int,
dropout: float = 0.4,
fc_dim: int = 256,
fine_tune_mode: str = "layer4",
):
super().__init__()
weights = models.ResNet18_Weights.DEFAULT
self.backbone = models.resnet18(weights=weights)
in_features = self.backbone.fc.in_features
# Freeze everything first
for param in self.backbone.parameters():
param.requires_grad = False
# Fine-tuning strategy
if fine_tune_mode == "frozen":
pass
elif fine_tune_mode == "layer4":
for param in self.backbone.layer4.parameters():
param.requires_grad = True
elif fine_tune_mode == "full":
for param in self.backbone.parameters():
param.requires_grad = True
else:
raise ValueError(f"Unsupported fine_tune_mode: {fine_tune_mode}")
self.backbone.fc = nn.Sequential(
nn.Dropout(dropout),
nn.Linear(in_features, fc_dim),
nn.ReLU(),
nn.Dropout(dropout),
nn.Linear(fc_dim, num_classes),
)
# Always train classifier head
for param in self.backbone.fc.parameters():
param.requires_grad = True
def forward(self, x):
return self.backbone(x) |