Comments (8)
From the configs, try replacing the structure from hitnet
to serial
. And please tell me the results you got
from hit.
Do you mean to change the STRUCTURE: "hitnet" to STRUCTURE: "serial"? However, an error occurs when the training is performed after modification:
Traceback (most recent call last):
File "train_net.py", line 255, in
main()
File "train_net.py", line 245, in main
args.no_head)
File "train_net.py", line 100, in train
mem_active,
File "/17106/zwx/HIT-master/hit/engine/trainer.py", line 59, in do_train
loss_dict, weight_dict, metric_dict, pooled_feature = model(slow_video, fast_video, boxes, objects, keypoints, mem_extras)
File "/root/anaconda3/envs/hit/lib/python3.7/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/17106/zwx/HIT-master/hit/modeling/detector/action_detector.py", line 20, in forward
result, detector_losses, loss_weight, detector_metrics = self.roi_heads(slow_features, fast_features, boxes, objects, keypoints, extras, part_forward)
File "/root/anaconda3/envs/hit/lib/python3.7/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/17106/zwx/HIT-master/hit/modeling/roi_heads/roi_heads_3d.py", line 12, in forward
result, loss_action, loss_weight, accuracy_action = self.action(slow_features, fast_features, boxes, objects, keypoints, extras, part_forward)
File "/root/anaconda3/envs/hit/lib/python3.7/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/17106/zwx/HIT-master/hit/modeling/roi_heads/action_head/action_head.py", line 44, in forward
x, x_pooled, x_objects, x_keypoints, x_pose = self.feature_extractor(slow_features, fast_features, proposals, objects, keypoints, extras, part_forward)
File "/root/anaconda3/envs/hit/lib/python3.7/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/17106/zwx/HIT-master/hit/modeling/roi_heads/action_head/roi_action_feature_extractor.py", line 145, in forward
ia_feature, res_person, res_object, res_keypoint = self.hit_structure(person_pooled, proposals, object_pooled, objects, hands_pooled, keypoints, memory_person, None, None, phase="rgb")
ValueError: too many values to unpack (expected 4)
The result shows an incorrect number of parameters
from hit.
The result before modification was:
VideoAP_ 0.5: 84.80
VideoAP_ 0.2: 86.40
from hit.
Yes, hitnet
to serial
as you did.
Thanks for reporting the numbers. Others have pointed out disparities of results for different runs (though I used seed). I will fix the serial
bug and get back to you.
from hit.
By the way, what is the frame mAP result for that run?
from hit.
frame mAP : 81.02
from hit.
Hello author, I found that the weights of some parameters were not loaded during the training process, will this affect the final training results and mAP calculation? If so, what is the cause? Is the weight file I loaded incomplete? Could the author give me some suggestions, thank you very much.
Here are some log messages from the training process:
2023-11-12 10:39:16,803 hit.utils.model_serialization INFO: backbone.slow.res_nl4.res_2.btnk.conv3.bn.weight loaded from backbone.slow.res_nl4.res_2.btnk.conv3.bn.weight of shape (2048,)
2023-11-12 10:39:16,803 hit.utils.model_serialization INFO: backbone.slow.res_nl4.res_2.btnk.conv3.conv.weight loaded from backbone.slow.res_nl4.res_2.btnk.conv3.conv.weight of shape (2048, 512, 1, 1, 1)
2023-11-12 10:39:16,803 hit.utils.model_serialization INFO: roi_heads.action.feature_extractor.fc1.bias will not be loaded.
2023-11-12 10:39:16,803 hit.utils.model_serialization INFO: roi_heads.action.feature_extractor.fc1.weight will not be loaded.
from hit.
In your case, weights not being loaded might be because you are training from an existing checkpoint.
I uploaded the pretrained model. Possibly due to the dataset being very small, different runs might give different results. I also tested different checkpoints (for JHMDB the model converges very fast then start overfitting).
from hit.
Related Issues (20)
- As model is missing so how should we run it HOT 2
- PreTrained model on Multi Sports Dataset HOT 2
- python setup.py bdist_wheel did not run successfully. HOT 1
- Problems encountered during training HOT 3
- Calculate frame mAP HOT 2
- get frame mAP error HOT 3
- Some questions about the results HOT 1
- No hit._custom_cuda_ext was found HOT 4
- ImportError HOT 3
- Distributed training problem HOT 1
- J-HMDB Model link is not working HOT 2
- Misror error in window on pip install-e HOT 1
- demo HOT 1
- Loss function HOT 2
- TypeError: conv3d() received an invalid combination of arguments error on training and inference HOT 1
- pip install -e . raise error in ROIAlign3d_cuda.cu (Windows) HOT 2
- ImportError HOT 1
- size mismatch for my video HOT 1
- Retrieve result_ava.csv for inference HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from hit.