# Copyright 2022 The Plenoptix Team. All rights reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. """ Code to train model. """ import logging import os import random from typing import Dict import torch import torch.distributed as dist import umsgpack from torch.nn.parallel import DistributedDataParallel as DDP from torchtyping import TensorType from pyrad.cameras.cameras import get_camera, get_intrinsics_from_intrinsics_matrix from pyrad.cameras.rays import RayBundle from pyrad.data.dataloader import EvalDataloader, setup_dataset_eval, setup_dataset_train from pyrad.graphs.base import setup_graph from pyrad.optimizers.optimizers import setup_optimizers from pyrad.utils import profiler, writer from pyrad.utils.config import Config from pyrad.utils.decorators import check_main_thread from pyrad.utils.writer import EventName, TimeWriter from pyrad.viewer.backend import vis_utils from pyrad.viewer.backend.utils import get_intrinsics_matrix_and_camera_to_world_h from pyrad.viewer.backend.visualizer import ViewerWindow, Visualizer logging.getLogger("PIL").setLevel(logging.WARNING) class Trainer: """Training class Args: config (Config): The configuration object. local_rank (int, optional): Local rank of the process. Defaults to 0. world_size (int, optional): World size of the process. Defaults to 1. """ def __init__(self, config: Config, local_rank: int = 0, world_size: int = 1): self.config = config self.local_rank = local_rank self.world_size = world_size self.device = "cpu" if world_size == 0 else f"cuda:{local_rank}" # dataset variables self.dataset_inputs_train = None self.dataloader_train = None self.dataloader_eval = None # model variables self.graph = None self.optimizers = None self.start_step = 0 # logging variables writer.setup_event_writers(config.logging, max_iter=config.trainer.max_num_iterations) profiler.setup_profiler(config.logging) self.vis = None if self.config.viewer.enable: window = ViewerWindow(zmq_url=self.config.viewer.zmq_url) self.vis = Visualizer(window=window) logging.info("Connected to viewer at %s", self.config.viewer.zmq_url) self.vis.delete() else: logging.info("Continuing without viewer.") def setup(self, test_mode=False): """Setup the Trainer by calling other setup functions.""" self.dataset_inputs_train, self.dataloader_train = setup_dataset_train(self.config.data, device=self.device) _, self.dataloader_eval = setup_dataset_eval(self.config.data, test_mode=test_mode, device=self.device) self.graph = setup_graph(self.config.graph, self.dataset_inputs_train, device=self.device) self.optimizers = setup_optimizers(self.config.optimizers, self.graph.get_param_groups()) if self.config.trainer.resume_train.load_dir: self._load_checkpoint() if self.world_size > 1: self.graph = DDP(self.graph, device_ids=[self.local_rank]) dist.barrier(device_ids=[self.local_rank]) self.graph.register_callbacks() @classmethod def get_aggregated_loss(cls, loss_dict: Dict[str, torch.tensor]): """Returns the aggregated losses and the scalar for calling .backwards() on. # TODO: move this out to another file/class/etc. """ loss_sum = 0.0 for loss_name in loss_dict.keys(): # TODO(ethan): add loss weightings here from a config loss_sum += loss_dict[loss_name] return loss_sum def train(self) -> None: """Train the model.""" if self.vis: self.draw_scene_in_viewer() with TimeWriter(writer, EventName.TOTAL_TRAIN_TIME): num_iterations = self.config.trainer.max_num_iterations iter_dataloader_train = iter(self.dataloader_train) for step in range(self.start_step, self.start_step + num_iterations): with TimeWriter(writer, EventName.ITER_LOAD_TIME, step=step): ray_indices, batch = next(iter_dataloader_train) with TimeWriter(writer, EventName.ITER_TRAIN_TIME, step=step) as t: loss_dict = self.train_iteration(ray_indices, batch, step) writer.put_scalar(name=EventName.RAYS_PER_SEC, scalar=ray_indices.shape[0] / t.duration, step=step) if step != 0 and step % self.config.logging.steps_per_log == 0: writer.put_dict(name="Loss/train-loss_dict", scalar_dict=loss_dict, step=step) if step != 0 and self.config.trainer.steps_per_save and step % self.config.trainer.steps_per_save == 0: self._save_checkpoint(self.config.trainer.model_dir, step) if ( self.vis and step != 0 and self.config.viewer.steps_per_render_image and step % self.config.viewer.steps_per_render_image == 0 ): _ = self.render_image_in_viewer() if step % self.config.trainer.steps_per_test == 0: self.eval_with_dataloader(self.dataloader_eval, step=step) self._write_out_storage(step) self._write_out_storage(num_iterations) def _write_out_storage(self, step: int) -> None: """Perform writes only during appropriate time steps Args: step (int): Current training step. """ if ( step % self.config.logging.steps_per_log == 0 or (self.config.trainer.steps_per_save and step % self.config.trainer.steps_per_save == 0) or step % self.config.trainer.steps_per_test == 0 or step == self.config.trainer.max_num_iterations ): writer.write_out_storage() def _load_checkpoint(self) -> None: """helper function to load graph and optimizer from prespecified checkpoint""" load_config = self.config.trainer.resume_train load_path = os.path.join(load_config.load_dir, f"step-{load_config.load_step:09d}.ckpt") assert os.path.exists(load_path), f"Checkpoint {load_path} does not exist" loaded_state = torch.load(load_path, map_location="cpu") self.start_step = loaded_state["step"] + 1 # load the checkpoints for graph and optimizer self.graph.load_graph(loaded_state) self.optimizers.load_optimizers(loaded_state) logging.info("done loading checkpoint from %s", load_path) @check_main_thread def _save_checkpoint(self, output_dir: str, step: int) -> None: """Save the model and optimizers Args: output_dir (str): directory to save the checkpoint step (int): number of steps in training for given checkpoint """ if not os.path.exists(output_dir): os.makedirs(output_dir) ckpt_path = os.path.join(output_dir, f"step-{step:09d}.ckpt") torch.save( { "step": step, "model": self.graph.module.state_dict() if hasattr(self.graph, "module") else self.graph.state_dict(), "optimizers": {k: v.state_dict() for (k, v) in self.optimizers.optimizers.items()}, }, ckpt_path, ) @profiler.time_function def train_iteration(self, ray_indices: TensorType["num_rays", 3], batch: dict, step: int) -> Dict[str, float]: """Run one iteration with a batch of inputs. Args: ray_indices (TensorType["num_rays", 3]): Contains camera, row, and col indicies for target rays. batch (dict): Batch of training data. step (int): Current training step. Returns: Dict[str, float]: Dictionary of model losses. """ _, loss_dict = self.graph.forward(ray_indices, batch=batch) loss = loss_dict["aggregated_loss"] self.optimizers.zero_grad_all() loss.backward() self.optimizers.optimizer_step_all() self.optimizers.scheduler_step_all(step) if self.graph.callbacks: for func_ in self.graph.callbacks: func_.after_step(step) return loss_dict @profiler.time_function def test_image(self, camera_ray_bundle: RayBundle, batch: dict, step: int = None) -> float: """Test a specific image. Args: camera_ray_bundle (RayBundle): Bundle of test rays. batch (dict): Batch of data. step (int): Current training step. Returns: float: PSNR """ self.graph.eval() image_idx = int(camera_ray_bundle.camera_indices[0, 0]) outputs = self.graph.get_outputs_for_camera_ray_bundle(camera_ray_bundle) psnr = self.graph.log_test_image_outputs(image_idx, step, batch, outputs) self.graph.train() return psnr def draw_scene_in_viewer(self): """Draw some images and the scene aabb in the viewer.""" image_dataset_train = self.dataloader_train.image_sampler.dataset dataset_inputs = self.dataset_inputs_train indices = random.sample(range(len(image_dataset_train)), k=10) for idx in indices: image = image_dataset_train[idx]["image"] camera = get_camera(dataset_inputs.intrinsics[idx], dataset_inputs.camera_to_world[idx], None) pose = camera.get_camera_to_world().double().numpy() K = camera.get_intrinsics_matrix().double().numpy() vis_utils.draw_camera_frustum( self.vis, image=(image.double().numpy() * 255.0), pose=pose, K=K, height=1.0, name=f"image_dataset_train/{idx:06d}", displayed_focal_length=0.5, realistic=False, ) aabb = dataset_inputs.scene_bounds.aabb vis_utils.draw_aabb(self.vis, aabb, name="dataset_inputs_train/scene_bounds/aabb") @profiler.time_function def render_image_in_viewer(self): """ Draw an image using the current camera pose from the viewer. The image is sent of a TCP connection and then uses WebRTC to send it to the viewer. """ data = self.vis["/Cameras/Main Camera"].get_object() message = umsgpack.unpackb(data) camera_object = message["object"]["object"] image_height = self.config.viewer.render_image_height intrinsics_matrix, camera_to_world_h = get_intrinsics_matrix_and_camera_to_world_h( camera_object, image_height=image_height ) camera_to_world = camera_to_world_h[:3, :] intrinsics = get_intrinsics_from_intrinsics_matrix(intrinsics_matrix) camera = get_camera(intrinsics, camera_to_world) camera_ray_bundle = camera.get_camera_ray_bundle(device=self.device) camera_ray_bundle.num_rays_per_chunk = self.config.viewer.num_rays_per_chunk self.graph.eval() outputs = self.graph.get_outputs_for_camera_ray_bundle(camera_ray_bundle) self.graph.train() # gross hack to get the image key, depending on which keys the graph uses rgb_key = "rgb" if "rgb" in outputs else "rgb_fine" # TODO: make it such that the TCP connection doesn't need float64 image = outputs[rgb_key].cpu().numpy().astype("float64") * 255 self.vis["/Cameras/Main Camera"].set_image(image) return outputs def eval_with_dataloader(self, dataloader: EvalDataloader, step: int = None) -> None: """Run evaluation with a given dataloader. Args: dataloader (EvalDataLoader): Evaluation dataloader. step (int): Current training iteration. """ for camera_ray_bundle, batch in dataloader: self.test_image(camera_ray_bundle, batch, step=step)