(nsr-pl) PS D:\Projects\PycharmProjects\instant-nsr-pl> python launch.py --config configs/nerf-colmap.yaml --gpu 0 --train dataset.scene=cup tag=example
Global seed set to 42
Using 16bit native Automatic Mixed Precision (AMP)
GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs
Trainer(limit_train_batches=1.0) was configured so 100% of the batches per epoch will be used..
Traceback (most recent call last):
File "D:\Projects\PycharmProjects\instant-nsr-pl\launch.py", line 123, in
main()
File "D:\Projects\PycharmProjects\instant-nsr-pl\launch.py", line 112, in main
trainer.fit(system, datamodule=dm)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 696, in fit
self._call_and_handle_interrupt(
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 650, in _call_and_handle_interrupt
return trainer_fn(*args, *kwargs)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 735, in _fit_impl
results = self._run(model, ckpt_path=self.ckpt_path)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1105, in _run
self._call_setup_hook() # allow user to setup lightning_module in accelerator environment
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1447, in _call_setup_hook
self._call_lightning_datamodule_hook("setup", stage=fn)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1569, in _call_lightning_datamodule_hook
return fn(args, **kwargs)
File "D:\Projects\PycharmProjects\instant-nsr-pl\datasets\colmap.py", line 235, in setup
self.train_dataset = ColmapIterableDataset(self.config, 'train')
File "D:\Projects\PycharmProjects\instant-nsr-pl\datasets\colmap.py", line 220, in init
self.setup(config, split)
File "D:\Projects\PycharmProjects\instant-nsr-pl\datasets\colmap.py", line 123, in setup
raise ValueError(f"Please parse the intrinsics for camera model {camdata[1].model}!")
raise ValueError(f"Please parse the intrinsics for camera model {camdata[1].model}!")
raise ValueError(f"Please parse the intrinsics for camera model {camdata[1].model}!")
ValueError: Please parse the intrinsics for camera model SIMPLE_PINHOLE!
Hello, may I ask after I made the data set with colmap, Run python launch.py --config configs/nerf-colmap.yaml --gpu 0 --train dataset.scene=cup tag=example. Please help me
(nsr-pl) PS D:\Projects\PycharmProjects\instant-nsr-pl> python launch.py --config configs/nerf-colmap.yaml --gpu 0 --train dataset.scene=cup tag=example
main()
File "D:\Projects\PycharmProjects\instant-nsr-pl\launch.py", line 112, in main
trainer.fit(system, datamodule=dm)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 696, in fit
self._call_and_handle_interrupt(
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 650, in _call_and_handle_interrupt
return trainer_fn(*args, *kwargs)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 735, in _fit_impl
results = self._run(model, ckpt_path=self.ckpt_path)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1105, in _run
self._call_setup_hook() # allow user to setup lightning_module in accelerator environment
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1447, in _call_setup_hook
self._call_lightning_datamodule_hook("setup", stage=fn)
File "D:\Anaconda\envs\nsr-pl\lib\site-packages\pytorch_lightning\trainer\trainer.py", line 1569, in _call_lightning_datamodule_hook
return fn(args, **kwargs)
File "D:\Projects\PycharmProjects\instant-nsr-pl\datasets\colmap.py", line 235, in setup
self.train_dataset = ColmapIterableDataset(self.config, 'train')
File "D:\Projects\PycharmProjects\instant-nsr-pl\datasets\colmap.py", line 220, in init
self.setup(config, split)
File "D:\Projects\PycharmProjects\instant-nsr-pl\datasets\colmap.py", line 123, in setup
raise ValueError(f"Please parse the intrinsics for camera model {camdata[1].model}!")
raise ValueError(f"Please parse the intrinsics for camera model {camdata[1].model}!")
raise ValueError(f"Please parse the intrinsics for camera model {camdata[1].model}!")
ValueError: Please parse the intrinsics for camera model SIMPLE_PINHOLE!
Global seed set to 42 Using 16bit native Automatic Mixed Precision (AMP) GPU available: True (cuda), used: True TPU available: False, using: 0 TPU cores IPU available: False, using: 0 IPUs HPU available: False, using: 0 HPUs
Trainer(limit_train_batches=1.0)
was configured so 100% of the batches per epoch will be used.. Traceback (most recent call last): File "D:\Projects\PycharmProjects\instant-nsr-pl\launch.py", line 123, inHello, may I ask after I made the data set with colmap, Run python launch.py --config configs/nerf-colmap.yaml --gpu 0 --train dataset.scene=cup tag=example. Please help me