refactoring configs; pulled network out of graph; created trainer heirarchy (#51)

* changed config graph structure

* removed network udner graphs; added new heirarchy trainer

* fixed callback in occupancy graph

* lint

* changed register callback to setter method
This commit is contained in:
evonneng authored and GitHub committed 2022-06-15 09:06:43 -07:00
1 parent 41073540c1
commit 91ef54963c
19 files changed
+163 -134

No files matched your search

+26 -15
View File
@@ -16,20 +16,31 @@
Callback functions for training iterations
"""
from pyrad.graphs.base import Graph
from pyrad.fields.occupancy_fields.occupancy_grid import OccupancyGrid
from typing import Callable
def update_occupancy(graph: Graph):
"""Update the occupancy grid depending on iteration count after every train step"""
if not hasattr(graph, "occupancy_pointers"):
occupancy_grids = []
for _, module in graph.__dict__["_modules"].items():
if isinstance(module, OccupancyGrid):
occupancy_grids.append(module)
graph.occupancy_pointers = occupancy_grids
for grid in graph.occupancy_pointers:
grid.iteration_count += 1
if grid.iteration_count % grid.update_every_num_iters == 0:
if grid.density_fn is not None:
grid.update_occupancy_grid(grid.density_fn)
class Callback:
"""Callback Object"""
def __init__(self, update_every_num_iters: int, func: Callable, *args, **kwargs):
self.update_every_num_iters = update_every_num_iters
self.func = func
self.args = args
self.kwargs = kwargs
def before_step(self, step: int):
"""callback to run before training step
Args:
step (int): current iteration step
"""
raise NotImplementedError
def after_step(self, step: int):
"""callback to run after training step
Args:
step (int): current iteration step
"""
if step % self.update_every_num_iters == 0:
self.func(*self.args, **self.kwargs)
+14 -14
View File
@@ -141,13 +141,13 @@ def setup_event_writers(config: DictConfig) -> None:
writer_class = getattr(sys.modules[__name__], writer_type)
writer_config = logging_configs[writer_type]
if writer_type == "LocalWriter":
curr_writer = writer_class(writer_config.save_dir, writer_config.stats_to_track, writer_config.max_log_size)
curr_writer = writer_class(writer_config.log_dir, writer_config.stats_to_track, writer_config.max_log_size)
else:
curr_writer = writer_class(writer_config.save_dir)
curr_writer = writer_class(writer_config.log_dir)
EVENT_WRITERS.append(curr_writer)
## configure all the global buffer basic information
GLOBAL_BUFFER["max_iter"] = config.graph.max_num_iterations
GLOBAL_BUFFER["max_iter"] = config.trainer.max_num_iterations
GLOBAL_BUFFER["max_buffer_size"] = config.logging.max_buffer_size
GLOBAL_BUFFER["steps_per_log"] = config.logging.steps_per_log
GLOBAL_BUFFER["events"] = {}
@@ -156,8 +156,8 @@ def setup_event_writers(config: DictConfig) -> None:
class Writer:
"""Writer class"""
def __init__(self, save_dir: str):
self.save_dir = save_dir
def __init__(self, log_dir: str):
self.log_dir = log_dir
@abstractmethod
def write_image(self, name: str, image: TensorType["H", "W", "C"], step: int) -> None:
@@ -225,9 +225,9 @@ class TimeWriter:
class WandbWriter(Writer):
"""WandDB Writer Class"""
def __init__(self, save_dir: str):
super().__init__(save_dir)
wandb.init(dir=save_dir)
def __init__(self, log_dir: str):
super().__init__(log_dir)
wandb.init(dir=log_dir)
def write_image(self, name: str, image: TensorType["H", "W", "C"], step: int) -> None:
"""_summary_
@@ -253,9 +253,9 @@ class WandbWriter(Writer):
class TensorboardWriter(Writer):
"""Tensorboard Writer Class"""
def __init__(self, save_dir: str):
super().__init__(save_dir)
self.tb_writer = SummaryWriter(log_dir=self.save_dir)
def __init__(self, log_dir: str):
super().__init__(log_dir)
self.tb_writer = SummaryWriter(log_dir=self.log_dir)
def write_image(self, name: str, image: TensorType["H", "W", "C"], step: int) -> None:
"""_summary_
@@ -310,13 +310,13 @@ def _format_time(seconds):
class LocalWriter(Writer):
"""Local Writer Class"""
def __init__(self, save_dir: str, stats_to_track: ListConfig, max_log_size: int = 0):
def __init__(self, log_dir: str, stats_to_track: ListConfig, max_log_size: int = 0):
"""
Args:
stats_to_track (ListConfig): the names of stats that should be logged.
max_log size (int): max number of lines that will be logged to teminal.
"""
super().__init__(save_dir)
super().__init__(log_dir)
self.stats_to_track = [EventName[name].value for name in stats_to_track]
self.max_log_size = max_log_size
self.keys = set()
@@ -325,7 +325,7 @@ class LocalWriter(Writer):
def write_image(self, name: str, image: TensorType["H", "W", "C"], step: int) -> None:
if name in self.stats_to_track:
image = to8b(image)
image_path = os.path.join(self.save_dir, f"{name}.jpg")
image_path = os.path.join(self.log_dir, f"{name}.jpg")
imageio.imwrite(image_path, np.uint8(image.cpu().numpy() * 255.0))
def write_scalar(self, name: str, scalar: float, step: int) -> None: