FX feature extraction

tests/test_models.py

@@ -4,9 +4,16 @@
 import os
 import fnmatch
 
+try:
+    from torchvision.models.feature_extraction import create_feature_extractor, get_graph_node_names, NodePathTracer
+    has_fx_feature_extraction = True
+except ImportError:
+    has_fx_feature_extraction = False
+
 import timm
 from timm import list_models, create_model, set_scriptable, has_model_default_key, is_model_default_key, \
     get_model_default_value
+from timm.models.fx_features import _leaf_modules, _autowrap_functions    
 
 if hasattr(torch._C, '_jit_set_profiling_executor'):
     # legacy executor is too slow to compile large models for unit tests
@@ -297,3 +304,145 @@ def test_model_forward_features(model_name, batch_size):
         assert e == o.shape[1]
         assert o.shape[0] == batch_size
         assert not torch.isnan(o).any()
+
+
+@pytest.mark.timeout(120)
+@pytest.mark.parametrize('model_name', list_models(exclude_filters=EXCLUDE_FILTERS))
+@pytest.mark.parametrize('batch_size', [1])
+def test_model_forward_fx(model_name, batch_size):
+    """
+    Symbolically trace each model and run single forward pass through the resulting GraphModule
+    Also check that the output of a forward pass through the GraphModule is the same as that from the original Module
+    """
+    if not has_fx_feature_extraction:
+        pytest.skip("Can't test FX because Torch >= 1.10 and Torchvision >= 0.11 are required")
+
+    model = create_model(model_name, pretrained=False)
+    model.eval()
+
+    input_size = _get_input_size(model=model, target=TARGET_FWD_SIZE)
+    if max(input_size) > MAX_FWD_SIZE:
+        pytest.skip("Fixed input size model > limit.")
+
+    # This block of code does a bit of juggling to handle any case where there are multiple outputs in train mode
+    # So we trace once and look at the graph, and get the indices of the nodes that lead into the original fx output
+    # node. Then we use those indices to select from train_nodes returned by torchvision get_graph_node_names
+    tracer = NodePathTracer(leaf_modules=list(_leaf_modules), autowrap_functions=list(_autowrap_functions))
+    graph = tracer.trace(model)
+    graph_nodes = list(reversed(graph.nodes))
+    output_node_names = [n.name for n in graph_nodes[0]._input_nodes.keys()]
+    graph_node_names = [n.name for n in graph_nodes]
+    output_node_indices = [-graph_node_names.index(node_name) for node_name in output_node_names]
+    train_nodes, eval_nodes = get_graph_node_names(
+        model, tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)})
+    eval_return_nodes = [eval_nodes[ix] for ix in output_node_indices]
+
+    fx_model = create_feature_extractor(
+        model, train_return_nodes=[train_nodes[-1]], eval_return_nodes=eval_return_nodes,
+        tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)})
+
+    inputs = torch.randn((batch_size, *input_size))
+    outputs = model(inputs)
+    if isinstance(outputs, tuple):
+        outputs = torch.cat(outputs)
+    fx_outputs = tuple(fx_model(inputs).values())
+    if isinstance(fx_outputs, tuple):
+        fx_outputs = torch.cat(fx_outputs)
+
+    assert torch.all(fx_outputs == outputs)
+    assert outputs.shape[0] == batch_size
+    assert not torch.isnan(outputs).any(), 'Output included NaNs'
+
+
+@pytest.mark.timeout(120)
+@pytest.mark.parametrize('model_name', list_models(exclude_filters=EXCLUDE_FILTERS, name_matches_cfg=True))
+@pytest.mark.parametrize('batch_size', [2])
+def test_model_backward_fx(model_name, batch_size):
+    """Symbolically trace each model and run single backward pass through the resulting GraphModule"""
+    if not has_fx_feature_extraction:
+        pytest.skip("Can't test FX because Torch >= 1.10 and Torchvision >= 0.11 are required")
+
+    input_size = _get_input_size(model_name=model_name, target=TARGET_BWD_SIZE)
+    if max(input_size) > MAX_BWD_SIZE:
+        pytest.skip("Fixed input size model > limit.")
+
+    model = create_model(model_name, pretrained=False, num_classes=42)
+    model.train()
+
+    num_params = sum([x.numel() for x in model.parameters()])
+
+    input_size = _get_input_size(model=model, target=TARGET_FWD_SIZE)
+    if max(input_size) > MAX_FWD_SIZE:
+        pytest.skip("Fixed input size model > limit.")
+
+    # This block of code does a bit of juggling to handle any case where there are multiple outputs in train mode
+    # So we trace once and look at the graph, and get the indices of the nodes that lead into the original fx output
+    # node. Then we use those indices to select from train_nodes returned by torchvision get_graph_node_names
+    tracer = NodePathTracer(leaf_modules=list(_leaf_modules), autowrap_functions=list(_autowrap_functions))
+    graph = tracer.trace(model)
+    graph_nodes = list(reversed(graph.nodes))
+    output_node_names = [n.name for n in graph_nodes[0]._input_nodes.keys()]
+    graph_node_names = [n.name for n in graph_nodes]
+    output_node_indices = [-graph_node_names.index(node_name) for node_name in output_node_names]
+    train_nodes, eval_nodes = get_graph_node_names(
+        model, tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)})
+    train_return_nodes = [train_nodes[ix] for ix in output_node_indices]
+
+    model = create_feature_extractor(
+        model, train_return_nodes=train_return_nodes, eval_return_nodes=[eval_nodes[-1]],
+        tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)})
+
+    inputs = torch.randn((batch_size, *input_size))
+    outputs = tuple(model(inputs).values())
+    if isinstance(outputs, tuple):
+        outputs = torch.cat(outputs)
+    outputs.mean().backward()
+    for n, x in model.named_parameters():
+        assert x.grad is not None, f'No gradient for {n}'
+    num_grad = sum([x.grad.numel() for x in model.parameters() if x.grad is not None])
+
+    assert outputs.shape[-1] == 42
+    assert num_params == num_grad, 'Some parameters are missing gradients'
+    assert not torch.isnan(outputs).any(), 'Output included NaNs'
+
+# reason: model is scripted after fx tracing, but beit has torch.jit.is_scripting() control flow
+EXCLUDE_FX_JIT_FILTERS = [
+    'beit_*',
+    'deit_*_distilled_patch16_224',
+    'levit*',
+    'pit_*_distilled_224',
+]
+
+@pytest.mark.timeout(120)
+@pytest.mark.parametrize(
+    'model_name', list_models(
+        exclude_filters=EXCLUDE_FILTERS + EXCLUDE_JIT_FILTERS + EXCLUDE_FX_JIT_FILTERS, name_matches_cfg=True))
+@pytest.mark.parametrize('batch_size', [1])
+def test_model_forward_fx_torchscript(model_name, batch_size):
+    """Symbolically trace each model, script it, and run single forward pass"""
+    if not has_fx_feature_extraction:
+        pytest.skip("Can't test FX because Torch >= 1.10 and Torchvision >= 0.11 are required")
+
+    input_size = _get_input_size(model_name=model_name, target=TARGET_JIT_SIZE)
+    if max(input_size) > MAX_JIT_SIZE:
+        pytest.skip("Fixed input size model > limit.")
+
+    with set_scriptable(True):
+        model = create_model(model_name, pretrained=False)
+    model.eval()
+
+    input_size = _get_input_size(model=model, target=TARGET_FWD_SIZE)
+    if max(input_size) > MAX_FWD_SIZE:
+        pytest.skip("Fixed input size model > limit.")
+
+    train_nodes, eval_nodes = get_graph_node_names(
+        model, tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)})
+    model = create_feature_extractor(
+        model, train_return_nodes=[train_nodes[-1]], eval_return_nodes=[eval_nodes[-1]],
+        tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)})
+
+    model = torch.jit.script(model)
+    outputs = model(torch.randn((batch_size, *input_size)))[train_nodes[-1]]
+
+    assert outputs.shape[0] == batch_size
+    assert not torch.isnan(outputs).any(), 'Output included NaNs'

timm/models/coat.py

@@ -19,6 +19,7 @@
 from .helpers import build_model_with_cfg, overlay_external_default_cfg
 from .layers import PatchEmbed, Mlp, DropPath, to_2tuple, trunc_normal_
 from .registry import register_model
+from .layers import _assert
 
 
 __all__ = [
@@ -105,7 +106,7 @@ def __init__(self, Ch, h, window):
     def forward(self, q, v, size: Tuple[int, int]):
         B, h, N, Ch = q.shape
         H, W = size
-        assert N == 1 + H * W
+        _assert(N == 1 + H * W, '')
 
         # Convolutional relative position encoding.
         q_img = q[:, :, 1:, :]  # [B, h, H*W, Ch]
@@ -177,7 +178,7 @@ def __init__(self, dim, k=3):
     def forward(self, x, size: Tuple[int, int]):
         B, N, C = x.shape
         H, W = size
-        assert N == 1 + H * W
+        _assert(N == 1 + H * W, '')
 
         # Extract CLS token and image tokens.
         cls_token, img_tokens = x[:, :1], x[:, 1:]  # [B, 1, C], [B, H*W, C]
@@ -275,7 +276,7 @@ def interpolate(self, x, scale_factor: float, size: Tuple[int, int]):
         """ Feature map interpolation. """
         B, N, C = x.shape
         H, W = size
-        assert N == 1 + H * W
+        _assert(N == 1 + H * W, '')
 
         cls_token = x[:, :1, :]
         img_tokens = x[:, 1:, :]

timm/models/convit.py

@@ -30,6 +30,7 @@
 from .layers import DropPath, to_2tuple, trunc_normal_, PatchEmbed, Mlp
 from .registry import register_model
 from .vision_transformer_hybrid import HybridEmbed
+from .fx_features import register_notrace_module
 
 import torch
 import torch.nn as nn
@@ -56,6 +57,7 @@ def _cfg(url='', **kwargs):
 }
 
 
+@register_notrace_module  # reason: FX can't symbolically trace control flow in forward method
 class GPSA(nn.Module):
     def __init__(self, dim, num_heads=8, qkv_bias=False, attn_drop=0., proj_drop=0.,
                  locality_strength=1.):

timm/models/crossvit.py

@@ -22,6 +22,7 @@
 Modifed from Timm. https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
 
 """
+from typing import Tuple
 
 import torch
 import torch.nn as nn
@@ -31,8 +32,9 @@
 from typing import List
 
 from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
+from .fx_features import register_notrace_function
 from .helpers import build_model_with_cfg
-from .layers import DropPath, to_2tuple, trunc_normal_
+from .layers import DropPath, to_2tuple, trunc_normal_, _assert
 from .registry import register_model
 from .vision_transformer import Mlp, Block
 
@@ -116,8 +118,10 @@ def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, multi
     def forward(self, x):
         B, C, H, W = x.shape
         # FIXME look at relaxing size constraints
-        assert H == self.img_size[0] and W == self.img_size[1], \
-            f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
+        _assert(H == self.img_size[0],
+                f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]}).")
+        _assert(W == self.img_size[1],
+                f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]}).")
         x = self.proj(x).flatten(2).transpose(1, 2)
         return x
 
@@ -255,6 +259,27 @@ def _compute_num_patches(img_size, patches):
     return [i[0] // p * i[1] // p for i, p in zip(img_size, patches)]
 
 
+@register_notrace_function
+def scale_image(x, ss: Tuple[int, int], crop_scale: bool = False):  # annotations for torchscript
+    """
+    Pulled out of CrossViT.forward_features to bury conditional logic in a leaf node for FX tracing.
+    Args:
+        x (Tensor): input image
+        ss (tuple[int, int]): height and width to scale to
+        crop_scale (bool): whether to crop instead of interpolate to achieve the desired scale. Defaults to False
+    Returns:
+        Tensor: the "scaled" image batch tensor
+    """
+    H, W = x.shape[-2:]
+    if H != ss[0] or W != ss[1]:
+        if crop_scale and ss[0] <= H and ss[1] <= W:
+            cu, cl = int(round((H - ss[0]) / 2.)), int(round((W - ss[1]) / 2.))
+            x = x[:, :, cu:cu + ss[0], cl:cl + ss[1]]
+        else:
+            x = torch.nn.functional.interpolate(x, size=ss, mode='bicubic', align_corners=False)
+    return x
+
+
 class CrossViT(nn.Module):
     """ Vision Transformer with support for patch or hybrid CNN input stage
     """
@@ -342,17 +367,12 @@ def reset_classifier(self, num_classes, global_pool=''):
              range(self.num_branches)])
 
     def forward_features(self, x):
-        B, C, H, W = x.shape
+        B = x.shape[0]
         xs = []
         for i, patch_embed in enumerate(self.patch_embed):
             x_ = x
             ss = self.img_size_scaled[i]
-            if H != ss[0] or W != ss[1]:
-                if self.crop_scale and ss[0] <= H and ss[1] <= W:
-                    cu, cl = int(round((H - ss[0]) / 2.)), int(round((W - ss[1]) / 2.))
-                    x_ = x_[:, :, cu:cu + ss[0], cl:cl + ss[1]]
-                else:
-                    x_ = torch.nn.functional.interpolate(x_, size=ss, mode='bicubic', align_corners=False)
+            x_ = scale_image(x_, ss, self.crop_scale)
             x_ = patch_embed(x_)
             cls_tokens = self.cls_token_0 if i == 0 else self.cls_token_1  # hard-coded for torch jit script
             cls_tokens = cls_tokens.expand(B, -1, -1)

timm/models/fx_features.py

@@ -0,0 +1,74 @@
+""" PyTorch FX Based Feature Extraction Helpers
+Using https://pytorch.org/vision/stable/feature_extraction.html
+"""
+from typing import Callable
+from torch import nn
+
+from .features import _get_feature_info
+
+try:
+    from torchvision.models.feature_extraction import create_feature_extractor
+    has_fx_feature_extraction = True
+except ImportError:
+    has_fx_feature_extraction = False
+
+# Layers we went to treat as leaf modules
+from .layers import Conv2dSame, ScaledStdConv2dSame, BatchNormAct2d, BlurPool2d, CondConv2d, StdConv2dSame, DropPath
+from .layers.non_local_attn import BilinearAttnTransform
+from .layers.pool2d_same import MaxPool2dSame, AvgPool2dSame
+
+# NOTE: By default, any modules from timm.models.layers that we want to treat as leaf modules go here
+# BUT modules from timm.models should use the registration mechanism below
+_leaf_modules = {
+    BatchNormAct2d,  # reason: flow control for jit scripting
+    BilinearAttnTransform,  # reason: flow control t <= 1
+    BlurPool2d,  # reason: TypeError: F.conv2d received Proxy in groups=x.shape[1]
+    # Reason: get_same_padding has a max which raises a control flow error
+    Conv2dSame, MaxPool2dSame,  ScaledStdConv2dSame, StdConv2dSame, AvgPool2dSame,
+    CondConv2d,  # reason: TypeError: F.conv2d received Proxy in groups=self.groups * B (because B = x.shape[0])
+    DropPath,  # reason: TypeError: rand recieved Proxy in `size` argument
+}
+
+try:
+    from .layers import InplaceAbn
+    _leaf_modules.add(InplaceAbn)
+except ImportError:
+    pass
+
+
+def register_notrace_module(module: nn.Module):
+    """
+    Any module not under timm.models.layers should get this decorator if we don't want to trace through it.
+    """
+    _leaf_modules.add(module)
+    return module
+
+
+# Functions we want to autowrap (treat them as leaves)
+_autowrap_functions = set()
+
+
+def register_notrace_function(func: Callable):
+    """
+    Decorator for functions which ought not to be traced through
+    """
+    _autowrap_functions.add(func)
+    return func
+
+
+class FeatureGraphNet(nn.Module):
+    def __init__(self, model, out_indices, out_map=None):
+        super().__init__()
+        assert has_fx_feature_extraction, 'Please update to PyTorch 1.10+, torchvision 0.11+ for FX feature extraction'
+        self.feature_info = _get_feature_info(model, out_indices)
+        if out_map is not None:
+            assert len(out_map) == len(out_indices)
+        return_nodes = {info['module']: out_map[i] if out_map is not None else info['module']
+                        for i, info in enumerate(self.feature_info) if i in out_indices}
+        self.graph_module = create_feature_extractor(
+            model, return_nodes,
+            tracer_kwargs={'leaf_modules': list(_leaf_modules), 'autowrap_functions': list(_autowrap_functions)})
+
+    def forward(self, x):
+        return list(self.graph_module(x).values())
+

timm/models/helpers.py

@@ -14,6 +14,7 @@
 
 
 from .features import FeatureListNet, FeatureDictNet, FeatureHookNet
+from .fx_features import FeatureGraphNet
 from .hub import has_hf_hub, download_cached_file, load_state_dict_from_hf, load_state_dict_from_url
 from .layers import Conv2dSame, Linear
 
@@ -477,6 +478,8 @@ def build_model_with_cfg(
                 feature_cls = feature_cls.lower()
                 if 'hook' in feature_cls:
                     feature_cls = FeatureHookNet
+                elif feature_cls == 'fx':
+                    feature_cls = FeatureGraphNet
                 else:
                     assert False, f'Unknown feature class {feature_cls}'
         model = feature_cls(model, **feature_cfg)