Model builder (#29)

Co-authored-by: Ayush Chaurasia <ayush.chuararsia@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

ultralytics/yolo/utils/torch_utils.py

@@ -1,11 +1,20 @@
+import math
 import os
+import time
 from contextlib import contextmanager
+from copy import deepcopy
+from pathlib import Path
 
+import thop
 import torch
 import torch.distributed as dist
+import torch.nn as nn
+import torch.nn.functional as F
 from torch.nn.parallel import DistributedDataParallel as DDP
 
-from ultralytics.yolo.utils import check_version
+from ultralytics.yolo.utils import LOGGER
+
+from .checks import check_version
 
 LOCAL_RANK = int(os.getenv('LOCAL_RANK', -1))  # https://pytorch.org/docs/stable/elastic/run.html
 RANK = int(os.getenv('RANK', -1))
@@ -68,3 +77,100 @@ def select_device(device='', batch_size=0, newline=True):
         s = s.rstrip()
     print(s)
     return torch.device(arg)
+
+
+def time_sync():
+    # PyTorch-accurate time
+    if torch.cuda.is_available():
+        torch.cuda.synchronize()
+    return time.time()
+
+
+def fuse_conv_and_bn(conv, bn):
+    # Fuse Conv2d() and BatchNorm2d() layers https://tehnokv.com/posts/fusing-batchnorm-and-conv/
+    fusedconv = nn.Conv2d(conv.in_channels,
+                          conv.out_channels,
+                          kernel_size=conv.kernel_size,
+                          stride=conv.stride,
+                          padding=conv.padding,
+                          dilation=conv.dilation,
+                          groups=conv.groups,
+                          bias=True).requires_grad_(False).to(conv.weight.device)
+
+    # Prepare filters
+    w_conv = conv.weight.clone().view(conv.out_channels, -1)
+    w_bn = torch.diag(bn.weight.div(torch.sqrt(bn.eps + bn.running_var)))
+    fusedconv.weight.copy_(torch.mm(w_bn, w_conv).view(fusedconv.weight.shape))
+
+    # Prepare spatial bias
+    b_conv = torch.zeros(conv.weight.size(0), device=conv.weight.device) if conv.bias is None else conv.bias
+    b_bn = bn.bias - bn.weight.mul(bn.running_mean).div(torch.sqrt(bn.running_var + bn.eps))
+    fusedconv.bias.copy_(torch.mm(w_bn, b_conv.reshape(-1, 1)).reshape(-1) + b_bn)
+
+    return fusedconv
+
+
+def model_info(model, verbose=False, imgsz=640):
+    # Model information. img_size may be int or list, i.e. img_size=640 or img_size=[640, 320]
+    n_p = sum(x.numel() for x in model.parameters())  # number parameters
+    n_g = sum(x.numel() for x in model.parameters() if x.requires_grad)  # number gradients
+    if verbose:
+        print(f"{'layer':>5} {'name':>40} {'gradient':>9} {'parameters':>12} {'shape':>20} {'mu':>10} {'sigma':>10}")
+        for i, (name, p) in enumerate(model.named_parameters()):
+            name = name.replace('module_list.', '')
+            print('%5g %40s %9s %12g %20s %10.3g %10.3g' %
+                  (i, name, p.requires_grad, p.numel(), list(p.shape), p.mean(), p.std()))
+
+    try:  # FLOPs
+        p = next(model.parameters())
+        stride = max(int(model.stride.max()), 32) if hasattr(model, 'stride') else 32  # max stride
+        im = torch.empty((1, p.shape[1], stride, stride), device=p.device)  # input image in BCHW format
+        flops = thop.profile(deepcopy(model), inputs=(im,), verbose=False)[0] / 1E9 * 2  # stride GFLOPs
+        imgsz = imgsz if isinstance(imgsz, list) else [imgsz, imgsz]  # expand if int/float
+        fs = f', {flops * imgsz[0] / stride * imgsz[1] / stride:.1f} GFLOPs'  # 640x640 GFLOPs
+    except Exception:
+        fs = ''
+
+    name = Path(model.yaml_file).stem.replace('yolov5', 'YOLOv5') if hasattr(model, 'yaml_file') else 'Model'
+    LOGGER.info(f"{name} summary: {len(list(model.modules()))} layers, {n_p} parameters, {n_g} gradients{fs}")
+
+
+def initialize_weights(model):
+    for m in model.modules():
+        t = type(m)
+        if t is nn.Conv2d:
+            pass  # nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
+        elif t is nn.BatchNorm2d:
+            m.eps = 1e-3
+            m.momentum = 0.03
+        elif t in [nn.Hardswish, nn.LeakyReLU, nn.ReLU, nn.ReLU6, nn.SiLU]:
+            m.inplace = True
+
+
+def scale_img(img, ratio=1.0, same_shape=False, gs=32):  # img(16,3,256,416)
+    # Scales img(bs,3,y,x) by ratio constrained to gs-multiple
+    if ratio == 1.0:
+        return img
+    h, w = img.shape[2:]
+    s = (int(h * ratio), int(w * ratio))  # new size
+    img = F.interpolate(img, size=s, mode='bilinear', align_corners=False)  # resize
+    if not same_shape:  # pad/crop img
+        h, w = (math.ceil(x * ratio / gs) * gs for x in (h, w))
+    return F.pad(img, [0, w - s[1], 0, h - s[0]], value=0.447)  # value = imagenet mean
+
+
+def copy_attr(a, b, include=(), exclude=()):
+    # Copy attributes from b to a, options to only include [...] and to exclude [...]
+    for k, v in b.__dict__.items():
+        if (len(include) and k not in include) or k.startswith('_') or k in exclude:
+            continue
+        else:
+            setattr(a, k, v)
+
+
+def smart_inference_mode(torch_1_9=check_version(torch.__version__, '1.9.0')):
+    # Applies torch.inference_mode() decorator if torch>=1.9.0 else torch.no_grad() decorator
+    def decorate(fn):
+        return (torch.inference_mode if torch_1_9 else torch.no_grad)()(fn)
+
+    return decorate