nerfies
nerfies copied to clipboard
Error while running colab v2 Typeerror stack trace included in model_utils
Including the stacktrace for the type
TypeError Traceback (most recent call last)
/usr/local/lib/python3.7/dist-packages/nerfies/models.py in construct_nerf(key, config, batch_size, appearance_ids, camera_ids, warp_ids, near, far, use_warp_jacobian, use_weights) 485 }, 486 init_rays_dict, --> 487 warp_extra=warp_extra)['params'] 488 489 return model, params
/usr/local/lib/python3.7/dist-packages/nerfies/models.py in call(self, rays_dict, warp_extra, metadata_encoded, use_warp, return_points, return_weights, return_warp_jacobian, deterministic) 346 metadata_encoded=metadata_encoded, 347 return_points=return_points, --> 348 return_weights=True) 349 out = {'coarse': coarse_ret} 350
/usr/local/lib/python3.7/dist-packages/nerfies/models.py in render_samples(self, level, points, z_vals, directions, viewdirs, metadata, warp_extra, use_warp, use_warp_jacobian, metadata_encoded, return_points, return_weights) 283 return_weights=return_weights, 284 use_white_background=self.use_white_background, --> 285 sample_at_infinity=self.use_sample_at_infinity)) 286 287 return out
/usr/local/lib/python3.7/dist-packages/nerfies/model_utils.py in volumetric_rendering(rgb, sigma, z_vals, dirs, use_white_background, sample_at_infinity, return_weights, eps) 105 dists = jnp.concatenate([ 106 z_vals[..., 1:] - z_vals[..., :-1], --> 107 jnp.broadcast_to([last_sample_z], z_vals[..., :1].shape) 108 ], -1) 109 dists = dists * jnp.linalg.norm(dirs[..., None, :], axis=-1)
/usr/local/lib/python3.7/dist-packages/jax/_src/numpy/util.py in _broadcast_to(arr, shape)
/usr/local/lib/python3.7/dist-packages/jax/_src/numpy/util.py in _check_arraylike(fun_name, *args)
TypeError: broadcast_to requires ndarray or scalar arguments, got <class 'list'> at position 0.
I am also caught with this error. Hope it to be fixed soon.