Fix config path for PyTorch Hub

[Lijun-Yu]

#36
The relative path in hubconf.py cannot be handled properly.

import torch
model = torch.hub.load('ultralytics/yolov5', 'yolov5s', pretrained=True, force_reload=True)

Downloading: "https://github.com/ultralytics/yolov5/archive/master.zip" to /home/lijun/.cache/torch/hub/master.zip
---------------------------------------------------------------------------
FileNotFoundError                         Traceback (most recent call last)
<ipython-input-2-a446dedfedeb> in <module>
      1 import torch
----> 2 model = torch.hub.load('ultralytics/yolov5', 'yolov5s', pretrained=True, force_reload=True)

~/environments/miniconda3/envs/actev/lib/python3.7/site-packages/torch/hub.py in load(github, model, *args, **kwargs)
    367     entry = _load_entry_from_hubconf(hub_module, model)
    368 
--> 369     model = entry(*args, **kwargs)
    370 
    371     sys.path.remove(repo_dir)

~/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py in yolov5s(pretrained, channels, classes)
     46         pytorch model
     47     """
---> 48     return create('yolov5s', pretrained, channels, classes)
     49 
     50 

~/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py in create(name, pretrained, channels, classes)
     25         pytorch model
     26     """
---> 27     model = Model('models/%s.yaml' % name, channels, classes)
     28     if pretrained:
     29         ckpt = '%s.pt' % name  # checkpoint filename

~/.cache/torch/hub/ultralytics_yolov5_master/models/yolo.py in __init__(self, model_cfg, ch, nc)
     51             self.md = model_cfg  # model dict
     52         else:  # is *.yaml
---> 53             with open(model_cfg) as f:
     54                 self.md = yaml.load(f, Loader=yaml.FullLoader)  # model dict
     55 

FileNotFoundError: [Errno 2] No such file or directory: 'models/yolov5s.yaml'

Changing to the absolute path should fix it.

import torch
model = torch.hub.load('Lijun-Yu/yolov5', 'yolov5s', pretrained=True, force_reload=True)

Downloading: "https://github.com/Lijun-Yu/yolov5/archive/master.zip" to /home/lijun/.cache/torch/hub/master.zip

              from  n    params  module                                  arguments                     
  0             -1  1      3520  models.common.Focus                     [3, 32, 3]                    
  1             -1  1     18560  models.common.Conv                      [32, 64, 3, 2]                
  2             -1  1     20672  models.common.Bottleneck                [64, 64]                      
  3             -1  1     73984  models.common.Conv                      [64, 128, 3, 2]               
  4             -1  1    161152  models.common.BottleneckCSP             [128, 128, 3]                 
  5             -1  1    295424  models.common.Conv                      [128, 256, 3, 2]              
  6             -1  1    641792  models.common.BottleneckCSP             [256, 256, 3]                 
  7             -1  1   1180672  models.common.Conv                      [256, 512, 3, 2]              
  8             -1  1    656896  models.common.SPP                       [512, 512, [5, 9, 13]]        
  9             -1  1   1905152  models.common.BottleneckCSP             [512, 512, 2]                 
 10             -1  1   1248768  models.common.BottleneckCSP             [512, 512, 1, False]          
 11             -1  1    130815  torch.nn.modules.conv.Conv2d            [512, 255, 1, 1]              
 12             -2  1         0  torch.nn.modules.upsampling.Upsample    [None, 2, 'nearest']          
 13        [-1, 6]  1         0  models.common.Concat                    [1]                           
 14             -1  1    197120  models.common.Conv                      [768, 256, 1, 1]              
 15             -1  1    313088  models.common.BottleneckCSP             [256, 256, 1, False]          
 16             -1  1     65535  torch.nn.modules.conv.Conv2d            [256, 255, 1, 1]              
 17             -2  1         0  torch.nn.modules.upsampling.Upsample    [None, 2, 'nearest']          
 18        [-1, 4]  1         0  models.common.Concat                    [1]                           
 19             -1  1     49408  models.common.Conv                      [384, 128, 1, 1]              
 20             -1  1     78720  models.common.BottleneckCSP             [128, 128, 1, False]          
 21             -1  1     32895  torch.nn.modules.conv.Conv2d            [128, 255, 1, 1]              
 22   [-1, 16, 11]  1         0  models.yolo.Detect                      [80, [[10, 13, 16, 30, 33, 23], [30, 61, 62, 45, 59, 119], [116, 90, 156, 198, 373, 326]]]
Model Summary: 165 layers, 7.07417e+06 parameters, 7.07417e+06 gradients

Downloading https://drive.google.com/uc?export=download&id=1R5T6rIyy3lLwgFXNms8whc-387H0tMQO as yolov5s.pt... Done (3.3s)

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Enhancement to model loading in hubconf.py.

📊 Key Changes

• Added Python's os module to enable path manipulations.
• Improved the way the model configuration file path is constructed, using os.path.join and os.path.dirname.
• Adjusted the model checkpoint loading to explicitly map to CPU memory, ensuring compatibility across environments that may not have a GPU.

🎯 Purpose & Impact

• These changes provide clearer, environment-independent paths that reduce the risk of file-path errors across different operating systems.
• The mapping of state dictionaries to CPU memory when loading pretrained models ensures users without access to a GPU can still load models seamlessly.
• Users can benefit from a more robust and error-proof model instantiation, making the YOLOv5 model easier to work with for both development and deployment, potentially broadening the user base to those with less technical expertise.

[glenn-jocher]

@Lijun-Yu thank you for the PR!

I've added a map_location to cpu after noticing an additional error on cpu.