Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Tensor维度不一致 #116

Open
JHB11Hinson opened this issue Sep 11, 2024 · 0 comments
Open

Tensor维度不一致 #116

JHB11Hinson opened this issue Sep 11, 2024 · 0 comments

Comments

@JHB11Hinson
Copy link

JHB11Hinson commented Sep 11, 2024

代码如下:
特征提取:

fet = FeatureExtractionTool(config='/data/huangjingwang/SCI/MMSA/cut/config2.json')

infile_path=os.path.abspath('/data/huangjingwang/SCI/MMSA/cut/test_output/output_video_002.mp4')
outfile_path=os.path.abspath('/data/huangjingwang/SCI/MMSA/cut/test_output/feature_get.pkl')
text_path=os.path.abspath("/data/huangjingwang/SCI/MMSA/cut/input.txt")
fet.run_single(in_file=infile_path, out_file=outfile_path,text="What is love")
模型训练:
config=get_config_regression('mtfn','sims')

config['custom_feature'] = '/data/huangjingwang/SCI/MMSA/dataset/SIMS/unaligned_39.pkl'
MMSA_run('mtfn', 'sims',
gpu_ids=[0],
config=config,
res_save_dir='/data/huangjingwang/SCI/MMSA/results/SIMS',
model_save_dir='/data/huangjingwang/SCI/MMSA/models/SMIS')

模型测试:
MMSA_test(config=config,
weights_path='/data/huangjingwang/SCI/MMSA/models/SMIS/mtfn-sims.pth',
feature_path='/data/huangjingwang/SCI/MMSA/cut/test_output/feature_get.pkl')

config2.json内容:
{
"audio": {
"tool": "librosa",
"sample_rate": 22050,
"features": {
"mfcc":{
"n_mfcc":20,
"htk": true
},
"zero_crossing_rate":{},
"chroma_cqt":{}
}
},
"video": {
"tool": "openface",
"fps": 30,
"ultiFace": {"enable": false,
"device": "cuda:0",
"facedetScale": 0.25,
"minTrack": 10,
"numFailedDet": 10,
"minFaceSize": 1,
"cropScale": 0.4},
"average_over": 1,
"args": {
"hogalign": false,
"simalign": false,
"nobadaligned": false,
"landmark_2D": true,
"landmark_3D": true,
"pdmparams": true,
"head_pose": true,
"action_units": true,
"gaze": true,
"tracked": false
}
},
"text": {
"model": "bert",
"device": "cpu",
"pretrained": "bert-base-chinese",
"args": {}
}
}

报错信息:Traceback (most recent call last):
File "/data/huangjingwang/SCI/MMSA/cut/test2.py", line 21, in
MMSA_test(config=config,
File "/data/huangjingwang/.conda/envs/hjw_nlp/lib/python3.11/site-packages/MMSA/run.py", line 350, in MMSA_test
output = model(text, audio, vision)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/data/huangjingwang/.conda/envs/hjw_nlp/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/data/huangjingwang/.conda/envs/hjw_nlp/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/data/huangjingwang/.conda/envs/hjw_nlp/lib/python3.11/site-packages/MMSA/models/AMIO.py", line 49, in forward
return self.Model(text_x, audio_x, video_x, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/data/huangjingwang/.conda/envs/hjw_nlp/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/data/huangjingwang/.conda/envs/hjw_nlp/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/data/huangjingwang/.conda/envs/hjw_nlp/lib/python3.11/site-packages/MMSA/models/multiTask/MTFN.py", line 185, in forward
_text_h = torch.cat((add_one, text_h), dim=1)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: Tensors must have same number of dimensions: got 2 and 1

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant