Is there any way to install it with the latest pytorch, pytorch3D on Win 11 + VS2019
xqyd opened this issue · comments
Thanks for the great job!
It appears to me that your guide of installation is for the old version of pytorch, pytorch3d, and neural_renderer. In my case, I had to run everything on Win11 + VS2019. I can figure out the installation of pytorch3d (latest) with cuda 11.x. But I was facing difficulty of installing the old neural_renderer. I've tried the recipe from https://github.com/lstcutong/neural_renderer_pytorch-win10, but get no luck. I guess it cannot work with cuda > 11.0 ?
If I follow your intallation guidelines, the neural_renderer can be compiled without any problem, but the pytorch3d 0.4.0 will pop up tons of errors that related with c++14, nvcc, ninja, etc., just like the one reported at iPERDance/iPERCore#12.
Anyway, is there a way to install the neural_renderer with the latest pytorch, pytorch3d? Please help me out!
Sorry to hear about that. In fact, our code is only tested on ubuntu 18.0. For pytorch3d, it should be compatible with cuda < 11.0, e.g. 10.1, 10.2. You can try to install pytorch3d with the old cuda version. Besides, the neural_renderer just prepares the reference image for the training of the network. So, I suppose that you can create two virtual environments, one for preparing data, and another for training the network. The reference image is generated by this function:
HF-Avatar/utils/data_generation.py
Line 285 in d656820
And the call function is here:
HF-Avatar/texture_generation.py
Line 383 in d656820
Thanks for the heads up. I have a lot of other codes depend on cuda 11.x so I kind of have to use newer env for your code. Anyway, after searching around, I've found a solution to modify the neural_render for compiling successful with latest pytorch (1.12.1 + cuda 11.3) and latest pytorch3d (0.7.0). (https://zhuanlan.zhihu.com/p/454832455 )
However, I'm still facing problem when running dynamic_offsets_runner.py. The output follows.
==========
C:\Users\000\anaconda3\envs\avatar2\python.exe D:/HF-Avatar-main/dynamic_offsets_runner.py
deal with Body2D_2040_499
496
pose_lr: 0.0005 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 0: body_joints_loss 0.0012308069271966815 face_joints_loss 0.000253313803113997 a_pose_prior_loss 0.1399410367012024 temporal_loss 6.803683936595917e-05
pose_lr: 0.00048666666666666666 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 20: body_joints_loss 0.0007467379909940064 face_joints_loss 3.77301184926182e-05 a_pose_prior_loss 0.13823160529136658 temporal_loss 2.145877806469798e-05
pose_lr: 0.00047333333333333336 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 40: body_joints_loss 0.0005183664616197348 face_joints_loss 2.0153687728452496e-05 a_pose_prior_loss 0.13675881922245026 temporal_loss 1.1691283361869864e-05
pose_lr: 0.00046 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 60: body_joints_loss 0.0004000658227596432 face_joints_loss 1.825666186050512e-05 a_pose_prior_loss 0.13528133928775787 temporal_loss 7.7563081504195e-06
pose_lr: 0.00044666666666666666 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 80: body_joints_loss 0.0003288703446742147 face_joints_loss 1.7135611415142193e-05 a_pose_prior_loss 0.13368374109268188 temporal_loss 5.8272316891816445e-06
pose_lr: 0.0004333333333333333 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 100: body_joints_loss 0.0002824698749464005 face_joints_loss 1.6413721823482774e-05 a_pose_prior_loss 0.1320117712020874 temporal_loss 4.643287866201717e-06
pose_lr: 0.00042 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 120: body_joints_loss 0.00025019669556058943 face_joints_loss 1.5870618881308474e-05 a_pose_prior_loss 0.13035769760608673 temporal_loss 3.852675035886932e-06
pose_lr: 0.00040666666666666667 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 140: body_joints_loss 0.0002265885705128312 face_joints_loss 1.5448920748895034e-05 a_pose_prior_loss 0.1287619024515152 temporal_loss 3.302234972579754e-06
pose_lr: 0.0003933333333333333 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 160: body_joints_loss 0.00020858774951193482 face_joints_loss 1.5115073438209947e-05 a_pose_prior_loss 0.12723977863788605 temporal_loss 2.9017382985330187e-06
pose_lr: 0.00038 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 180: body_joints_loss 0.00019436916045378894 face_joints_loss 1.4846345948171802e-05 a_pose_prior_loss 0.1257980763912201 temporal_loss 2.601080495878705e-06
pose_lr: 0.00036666666666666667 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 200: body_joints_loss 0.00018279688083566725 face_joints_loss 1.4627429663960356e-05 a_pose_prior_loss 0.12443770468235016 temporal_loss 2.3696459265920566e-06
pose_lr: 0.0003533333333333333 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 220: body_joints_loss 0.00017314343131147325 face_joints_loss 1.4447070498135872e-05 a_pose_prior_loss 0.12315767258405685 temporal_loss 2.187597146985354e-06
pose_lr: 0.00034 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 240: body_joints_loss 0.00016492829308845103 face_joints_loss 1.4296879271569196e-05 a_pose_prior_loss 0.12195603549480438 temporal_loss 2.041735569946468e-06
pose_lr: 0.0003266666666666667 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 260: body_joints_loss 0.000157824601046741 face_joints_loss 1.417050134477904e-05 a_pose_prior_loss 0.12083037197589874 temporal_loss 1.9229919416829944e-06
pose_lr: 0.0003133333333333333 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 280: body_joints_loss 0.0001516036718385294 face_joints_loss 1.4063104572414886e-05 a_pose_prior_loss 0.11977805942296982 temporal_loss 1.8250307221023832e-06
pose_lr: 0.00030000000000000003 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 300: body_joints_loss 0.0001461013889638707 face_joints_loss 1.397097366861999e-05 a_pose_prior_loss 0.11879623681306839 temporal_loss 1.7432836330044665e-06
pose_lr: 0.0002866666666666667 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 320: body_joints_loss 0.00014119618572294712 face_joints_loss 1.3891319213144016e-05 a_pose_prior_loss 0.11788195371627808 temporal_loss 1.674360873948899e-06
pose_lr: 0.00027333333333333333 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 340: body_joints_loss 0.00013679714174941182 face_joints_loss 1.3821836546412669e-05 a_pose_prior_loss 0.11703215539455414 temporal_loss 1.6157628124346957e-06
pose_lr: 0.00026000000000000003 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 360: body_joints_loss 0.00013283314183354378 face_joints_loss 1.3760813999397215e-05 a_pose_prior_loss 0.11624372005462646 temporal_loss 1.5656100913474802e-06
pose_lr: 0.0002466666666666667 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 380: body_joints_loss 0.00012924824841320515 face_joints_loss 1.370685549773043e-05 a_pose_prior_loss 0.11551357060670853 temporal_loss 1.5224078424580512e-06
pose_lr: 0.00023333333333333333 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 400: body_joints_loss 0.00012599707406479865 face_joints_loss 1.3658857824339066e-05 a_pose_prior_loss 0.1148386225104332 temporal_loss 1.4849978242637007e-06
pose_lr: 0.00021999999999999998 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 420: body_joints_loss 0.00012304254050832242 face_joints_loss 1.3615909665531944e-05 a_pose_prior_loss 0.11421594768762589 temporal_loss 1.4523940308208694e-06
pose_lr: 0.00020666666666666668 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 440: body_joints_loss 0.0001203529755002819 face_joints_loss 1.357728979201056e-05 a_pose_prior_loss 0.11364272236824036 temporal_loss 1.4240174550650408e-06
pose_lr: 0.00019333333333333333 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 460: body_joints_loss 0.00011790287680923939 face_joints_loss 1.3542321539716795e-05 a_pose_prior_loss 0.1131162941455841 temporal_loss 1.3991066225571558e-06
pose_lr: 0.00017999999999999998 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 480: body_joints_loss 0.00011566960165509954 face_joints_loss 1.351050832454348e-05 a_pose_prior_loss 0.11263421177864075 temporal_loss 1.3772545344181708e-06
pose_lr: 0.00016666666666666663 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 500: body_joints_loss 0.00011363394150976092 face_joints_loss 1.3481453606800642e-05 a_pose_prior_loss 0.1121942549943924 temporal_loss 1.3580445283878362e-06
pose_lr: 0.00015333333333333334 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 520: body_joints_loss 0.00011177945998497307 face_joints_loss 1.345477176073473e-05 a_pose_prior_loss 0.1117943525314331 temporal_loss 1.3411510053629172e-06
pose_lr: 0.00014 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 540: body_joints_loss 0.00011009222362190485 face_joints_loss 1.3430152648652438e-05 a_pose_prior_loss 0.11143271625041962 temporal_loss 1.3263183973322157e-06
pose_lr: 0.00012666666666666664 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 560: body_joints_loss 0.00010855973232537508 face_joints_loss 1.3407360711426008e-05 a_pose_prior_loss 0.1111077293753624 temporal_loss 1.3132826097717043e-06
pose_lr: 0.00011333333333333334 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 580: body_joints_loss 0.00010717236727941781 face_joints_loss 1.3386098544287961e-05 a_pose_prior_loss 0.11081802099943161 temporal_loss 1.3019565585636883e-06
pose_lr: 9.999999999999999e-05 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 600: body_joints_loss 0.00010592067701509222 face_joints_loss 1.3366285202209838e-05 a_pose_prior_loss 0.11056239157915115 temporal_loss 1.2920801282234606e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 620: body_joints_loss 0.00010475726594449952 face_joints_loss 1.3347233107197098e-05 a_pose_prior_loss 0.11032479256391525 temporal_loss 1.28311148728244e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 640: body_joints_loss 0.00010363145702285692 face_joints_loss 1.3328671229828615e-05 a_pose_prior_loss 0.11008751392364502 temporal_loss 1.2743502111334237e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 660: body_joints_loss 0.00010254091466777027 face_joints_loss 1.3310573194758035e-05 a_pose_prior_loss 0.10984986275434494 temporal_loss 1.265866785615799e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 680: body_joints_loss 0.00010148392902920023 face_joints_loss 1.3292914445628412e-05 a_pose_prior_loss 0.10961173474788666 temporal_loss 1.2576028893818147e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 700: body_joints_loss 0.00010045933595392853 face_joints_loss 1.3275616765895393e-05 a_pose_prior_loss 0.10937319695949554 temporal_loss 1.2495664805101114e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 720: body_joints_loss 9.946515638148412e-05 face_joints_loss 1.3258785656944383e-05 a_pose_prior_loss 0.10913429409265518 temporal_loss 1.2417086736604688e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 740: body_joints_loss 9.850115020526573e-05 face_joints_loss 1.3242164641269483e-05 a_pose_prior_loss 0.10889510810375214 temporal_loss 1.233995476468408e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 760: body_joints_loss 9.756530926097184e-05 face_joints_loss 1.3225880138634238e-05 a_pose_prior_loss 0.10865574330091476 temporal_loss 1.22650169487315e-06
pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001
epoch_idx 780: body_joints_loss 9.665657125879079e-05 face_joints_loss 1.3209883036324754e-05 a_pose_prior_loss 0.10841624438762665 temporal_loss 1.2192133453936549e-06
100%|██████████| 496/496 [11:06<00:00, 1.34s/it]
--------------- Epoch: 0 ---------------
C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\torch\nn\functional.py:4215: UserWarning: Default grid_sample and affine_grid behavior has changed to align_corners=False since 1.3.0. Please specify align_corners=True if the old behavior is desired. See the documentation of grid_sample for details.
warnings.warn(
C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\transforms\transform3d.py:802: UserWarning: R is not a valid rotation matrix
warnings.warn(msg)
Traceback (most recent call last):
File "D:/HF-Avatar-main/dynamic_offsets_runner.py", line 345, in
trainer.train()
File "D:/HF-Avatar-main/dynamic_offsets_runner.py", line 177, in train
loss = self.forward(step, item, input, ground_truth)
File "D:/HF-Avatar-main/dynamic_offsets_runner.py", line 144, in forward
model_output = self.network.forward(item, input['naked_vertice_uv'])
File "D:\HF-Avatar-main\geometry_model\network.py", line 815, in forward
images = self.renderer(mesh)
File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\renderer.py", line 62, in forward
images = self.shader(fragments, meshes_world, **kwargs)
File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\shader.py", line 126, in forward
texels = meshes.sample_textures(fragments)
File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\structures\meshes.py", line 1552, in sample_textures
return self.textures.sample_textures(
File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\textures.py", line 958, in sample_textures
packing_list = [
File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\textures.py", line 959, in
i[j] for i, j in zip(self.verts_uvs_list(), self.faces_uvs_list())
IndexError: tensors used as indices must be long, byte or bool tensors
Process finished with exit code 1
Any idea how to solve it? Thanks!
Hard to see, it seems that the face index is not loaded correctly. Please check this line:
HF-Avatar/geometry_model/network.py
Line 688 in d656820
f.
Thanks for the heads up. I have a lot of other codes depend on cuda 11.x so I kind of have to use newer env for your code. Anyway, after searching around, I've found a solution to modify the neural_render for compiling successful with latest pytorch (1.12.1 + cuda 11.3) and latest pytorch3d (0.7.0). (https://zhuanlan.zhihu.com/p/454832455 )
However, I'm still facing problem when running dynamic_offsets_runner.py. The output follows.
========== C:\Users\000\anaconda3\envs\avatar2\python.exe D:/HF-Avatar-main/dynamic_offsets_runner.py deal with Body2D_2040_499 496 pose_lr: 0.0005 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 0: body_joints_loss 0.0012308069271966815 face_joints_loss 0.000253313803113997 a_pose_prior_loss 0.1399410367012024 temporal_loss 6.803683936595917e-05 pose_lr: 0.00048666666666666666 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 20: body_joints_loss 0.0007467379909940064 face_joints_loss 3.77301184926182e-05 a_pose_prior_loss 0.13823160529136658 temporal_loss 2.145877806469798e-05 pose_lr: 0.00047333333333333336 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 40: body_joints_loss 0.0005183664616197348 face_joints_loss 2.0153687728452496e-05 a_pose_prior_loss 0.13675881922245026 temporal_loss 1.1691283361869864e-05 pose_lr: 0.00046 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 60: body_joints_loss 0.0004000658227596432 face_joints_loss 1.825666186050512e-05 a_pose_prior_loss 0.13528133928775787 temporal_loss 7.7563081504195e-06 pose_lr: 0.00044666666666666666 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 80: body_joints_loss 0.0003288703446742147 face_joints_loss 1.7135611415142193e-05 a_pose_prior_loss 0.13368374109268188 temporal_loss 5.8272316891816445e-06 pose_lr: 0.0004333333333333333 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 100: body_joints_loss 0.0002824698749464005 face_joints_loss 1.6413721823482774e-05 a_pose_prior_loss 0.1320117712020874 temporal_loss 4.643287866201717e-06 pose_lr: 0.00042 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 120: body_joints_loss 0.00025019669556058943 face_joints_loss 1.5870618881308474e-05 a_pose_prior_loss 0.13035769760608673 temporal_loss 3.852675035886932e-06 pose_lr: 0.00040666666666666667 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 140: body_joints_loss 0.0002265885705128312 face_joints_loss 1.5448920748895034e-05 a_pose_prior_loss 0.1287619024515152 temporal_loss 3.302234972579754e-06 pose_lr: 0.0003933333333333333 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 160: body_joints_loss 0.00020858774951193482 face_joints_loss 1.5115073438209947e-05 a_pose_prior_loss 0.12723977863788605 temporal_loss 2.9017382985330187e-06 pose_lr: 0.00038 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 180: body_joints_loss 0.00019436916045378894 face_joints_loss 1.4846345948171802e-05 a_pose_prior_loss 0.1257980763912201 temporal_loss 2.601080495878705e-06 pose_lr: 0.00036666666666666667 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 200: body_joints_loss 0.00018279688083566725 face_joints_loss 1.4627429663960356e-05 a_pose_prior_loss 0.12443770468235016 temporal_loss 2.3696459265920566e-06 pose_lr: 0.0003533333333333333 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 220: body_joints_loss 0.00017314343131147325 face_joints_loss 1.4447070498135872e-05 a_pose_prior_loss 0.12315767258405685 temporal_loss 2.187597146985354e-06 pose_lr: 0.00034 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 240: body_joints_loss 0.00016492829308845103 face_joints_loss 1.4296879271569196e-05 a_pose_prior_loss 0.12195603549480438 temporal_loss 2.041735569946468e-06 pose_lr: 0.0003266666666666667 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 260: body_joints_loss 0.000157824601046741 face_joints_loss 1.417050134477904e-05 a_pose_prior_loss 0.12083037197589874 temporal_loss 1.9229919416829944e-06 pose_lr: 0.0003133333333333333 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 280: body_joints_loss 0.0001516036718385294 face_joints_loss 1.4063104572414886e-05 a_pose_prior_loss 0.11977805942296982 temporal_loss 1.8250307221023832e-06 pose_lr: 0.00030000000000000003 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 300: body_joints_loss 0.0001461013889638707 face_joints_loss 1.397097366861999e-05 a_pose_prior_loss 0.11879623681306839 temporal_loss 1.7432836330044665e-06 pose_lr: 0.0002866666666666667 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 320: body_joints_loss 0.00014119618572294712 face_joints_loss 1.3891319213144016e-05 a_pose_prior_loss 0.11788195371627808 temporal_loss 1.674360873948899e-06 pose_lr: 0.00027333333333333333 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 340: body_joints_loss 0.00013679714174941182 face_joints_loss 1.3821836546412669e-05 a_pose_prior_loss 0.11703215539455414 temporal_loss 1.6157628124346957e-06 pose_lr: 0.00026000000000000003 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 360: body_joints_loss 0.00013283314183354378 face_joints_loss 1.3760813999397215e-05 a_pose_prior_loss 0.11624372005462646 temporal_loss 1.5656100913474802e-06 pose_lr: 0.0002466666666666667 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 380: body_joints_loss 0.00012924824841320515 face_joints_loss 1.370685549773043e-05 a_pose_prior_loss 0.11551357060670853 temporal_loss 1.5224078424580512e-06 pose_lr: 0.00023333333333333333 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 400: body_joints_loss 0.00012599707406479865 face_joints_loss 1.3658857824339066e-05 a_pose_prior_loss 0.1148386225104332 temporal_loss 1.4849978242637007e-06 pose_lr: 0.00021999999999999998 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 420: body_joints_loss 0.00012304254050832242 face_joints_loss 1.3615909665531944e-05 a_pose_prior_loss 0.11421594768762589 temporal_loss 1.4523940308208694e-06 pose_lr: 0.00020666666666666668 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 440: body_joints_loss 0.0001203529755002819 face_joints_loss 1.357728979201056e-05 a_pose_prior_loss 0.11364272236824036 temporal_loss 1.4240174550650408e-06 pose_lr: 0.00019333333333333333 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 460: body_joints_loss 0.00011790287680923939 face_joints_loss 1.3542321539716795e-05 a_pose_prior_loss 0.1131162941455841 temporal_loss 1.3991066225571558e-06 pose_lr: 0.00017999999999999998 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 480: body_joints_loss 0.00011566960165509954 face_joints_loss 1.351050832454348e-05 a_pose_prior_loss 0.11263421177864075 temporal_loss 1.3772545344181708e-06 pose_lr: 0.00016666666666666663 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 500: body_joints_loss 0.00011363394150976092 face_joints_loss 1.3481453606800642e-05 a_pose_prior_loss 0.1121942549943924 temporal_loss 1.3580445283878362e-06 pose_lr: 0.00015333333333333334 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 520: body_joints_loss 0.00011177945998497307 face_joints_loss 1.345477176073473e-05 a_pose_prior_loss 0.1117943525314331 temporal_loss 1.3411510053629172e-06 pose_lr: 0.00014 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 540: body_joints_loss 0.00011009222362190485 face_joints_loss 1.3430152648652438e-05 a_pose_prior_loss 0.11143271625041962 temporal_loss 1.3263183973322157e-06 pose_lr: 0.00012666666666666664 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 560: body_joints_loss 0.00010855973232537508 face_joints_loss 1.3407360711426008e-05 a_pose_prior_loss 0.1111077293753624 temporal_loss 1.3132826097717043e-06 pose_lr: 0.00011333333333333334 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 580: body_joints_loss 0.00010717236727941781 face_joints_loss 1.3386098544287961e-05 a_pose_prior_loss 0.11081802099943161 temporal_loss 1.3019565585636883e-06 pose_lr: 9.999999999999999e-05 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 600: body_joints_loss 0.00010592067701509222 face_joints_loss 1.3366285202209838e-05 a_pose_prior_loss 0.11056239157915115 temporal_loss 1.2920801282234606e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 620: body_joints_loss 0.00010475726594449952 face_joints_loss 1.3347233107197098e-05 a_pose_prior_loss 0.11032479256391525 temporal_loss 1.28311148728244e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 640: body_joints_loss 0.00010363145702285692 face_joints_loss 1.3328671229828615e-05 a_pose_prior_loss 0.11008751392364502 temporal_loss 1.2743502111334237e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 660: body_joints_loss 0.00010254091466777027 face_joints_loss 1.3310573194758035e-05 a_pose_prior_loss 0.10984986275434494 temporal_loss 1.265866785615799e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 680: body_joints_loss 0.00010148392902920023 face_joints_loss 1.3292914445628412e-05 a_pose_prior_loss 0.10961173474788666 temporal_loss 1.2576028893818147e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 700: body_joints_loss 0.00010045933595392853 face_joints_loss 1.3275616765895393e-05 a_pose_prior_loss 0.10937319695949554 temporal_loss 1.2495664805101114e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 720: body_joints_loss 9.946515638148412e-05 face_joints_loss 1.3258785656944383e-05 a_pose_prior_loss 0.10913429409265518 temporal_loss 1.2417086736604688e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 740: body_joints_loss 9.850115020526573e-05 face_joints_loss 1.3242164641269483e-05 a_pose_prior_loss 0.10889510810375214 temporal_loss 1.233995476468408e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 760: body_joints_loss 9.756530926097184e-05 face_joints_loss 1.3225880138634238e-05 a_pose_prior_loss 0.10865574330091476 temporal_loss 1.22650169487315e-06 pose_lr: 0.0001 betas_lr: 0.001 trans_lr: 0.001 epoch_idx 780: body_joints_loss 9.665657125879079e-05 face_joints_loss 1.3209883036324754e-05 a_pose_prior_loss 0.10841624438762665 temporal_loss 1.2192133453936549e-06 100%|██████████| 496/496 [11:06<00:00, 1.34s/it] --------------- Epoch: 0 --------------- C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\torch\nn\functional.py:4215: UserWarning: Default grid_sample and affine_grid behavior has changed to align_corners=False since 1.3.0. Please specify align_corners=True if the old behavior is desired. See the documentation of grid_sample for details. warnings.warn( C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\transforms\transform3d.py:802: UserWarning: R is not a valid rotation matrix warnings.warn(msg) Traceback (most recent call last): File "D:/HF-Avatar-main/dynamic_offsets_runner.py", line 345, in trainer.train() File "D:/HF-Avatar-main/dynamic_offsets_runner.py", line 177, in train loss = self.forward(step, item, input, ground_truth) File "D:/HF-Avatar-main/dynamic_offsets_runner.py", line 144, in forward model_output = self.network.forward(item, input['naked_vertice_uv']) File "D:\HF-Avatar-main\geometry_model\network.py", line 815, in forward images = self.renderer(mesh) File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\renderer.py", line 62, in forward images = self.shader(fragments, meshes_world, **kwargs) File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\shader.py", line 126, in forward texels = meshes.sample_textures(fragments) File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\structures\meshes.py", line 1552, in sample_textures return self.textures.sample_textures( File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\textures.py", line 958, in sample_textures packing_list = [ File "C:\Users\000\anaconda3\envs\avatar2\lib\site-packages\pytorch3d-0.7.0-py3.8-win-amd64.egg\pytorch3d\renderer\mesh\textures.py", line 959, in i[j] for i, j in zip(self.verts_uvs_list(), self.faces_uvs_list()) IndexError: tensors used as indices must be long, byte or bool tensors
Process finished with exit code 1
Any idea how to solve it? Thanks!
I encountered the same problem, how did you solve it? thanks
Hard to see, it seems that the face index is not loaded correctly. Please check this line:
HF-Avatar/geometry_model/network.py
Line 688 in d656820
the value of
f.
Hello, how can I solve it?Thanks
Hard to see, it seems that the face index is not loaded correctly. Please check this line:
HF-Avatar/geometry_model/network.py
Line 688 in d656820
the value of
f.Hello, how can I solve it?Thanks
print(f) directly.
In the meanwhile, we only test the code with pytorch3d=0.4.0. There is no guarantee that other versions, like 0.7.0, will work.
Hi did you solve it ? I have many error such as pytorch3d and neural-renderer-pytorch. Can someone help me to solve it ?