make visdom optional

Summary: Make Implicitron run without visdom installed.

Reviewed By: shapovalov

Differential Revision: D40587974

fbshipit-source-id: dc319596c7a4d10a4c54c556dabc89ad9d25c2fb

Author: bottler

projects/implicitron_trainer/experiment.py

@@ -41,7 +41,7 @@
         Stats are logged and plotted to the file "train_stats.pdf" in the
         same directory. The stats are also saved as part of the checkpoint file.
   - Visualizations
-        Prredictions are plotted to a visdom server running at the
+        Predictions are plotted to a visdom server running at the
         port specified by the `visdom_server` and `visdom_port` keys in the
         config file.
 

pytorch3d/implicitron/evaluation/evaluate_new_view_synthesis.py

@@ -9,7 +9,7 @@
 import warnings
 from collections import OrderedDict
 from dataclasses import dataclass, field
-from typing import Any, Dict, List, Optional, Sequence, Tuple, Union
+from typing import Any, Dict, List, Optional, Sequence, Tuple, TYPE_CHECKING, Union
 
 import numpy as np
 import torch
@@ -27,7 +27,9 @@
 from pytorch3d.renderer.cameras import CamerasBase, PerspectiveCameras
 from pytorch3d.vis.plotly_vis import plot_scene
 from tabulate import tabulate
-from visdom import Visdom
+
+if TYPE_CHECKING:
+    from visdom import Visdom
 
 
 EVAL_N_SRC_VIEWS = [1, 3, 5, 7, 9]
@@ -43,14 +45,16 @@ class _Visualizer:
 
     visdom_env: str = "eval_debug"
 
-    _viz: Visdom = field(init=False)
+    _viz: Optional["Visdom"] = field(init=False)
 
     def __post_init__(self):
         self._viz = vis_utils.get_visdom_connection()
 
     def show_rgb(
         self, loss_value: float, metric_name: str, loss_mask_now: torch.Tensor
     ):
+        if self._viz is None:
+            return
         self._viz.images(
             torch.cat(
                 (
@@ -68,7 +72,10 @@ def show_rgb(
     def show_depth(
         self, depth_loss: float, name_postfix: str, loss_mask_now: torch.Tensor
     ):
-        self._viz.images(
+        if self._viz is None:
+            return
+        viz = self._viz
+        viz.images(
             torch.cat(
                 (
                     make_depth_image(self.depth_render, loss_mask_now),
@@ -80,13 +87,13 @@ def show_depth(
             win="depth_abs" + name_postfix,
             opts={"title": f"depth_abs_{name_postfix}_{depth_loss:1.2f}"},
         )
-        self._viz.images(
+        viz.images(
             loss_mask_now,
             env=self.visdom_env,
             win="depth_abs" + name_postfix + "_mask",
             opts={"title": f"depth_abs_{name_postfix}_{depth_loss:1.2f}_mask"},
         )
-        self._viz.images(
+        viz.images(
             self.depth_mask,
             env=self.visdom_env,
             win="depth_abs" + name_postfix + "_maskd",
@@ -126,7 +133,7 @@ def show_depth(
             pointcloud_max_points=10000,
             pointcloud_marker_size=1,
         )
-        self._viz.plotlyplot(
+        viz.plotlyplot(
             plotlyplot,
             env=self.visdom_env,
             win=f"pcl{name_postfix}",

pytorch3d/implicitron/models/generic_model.py

@@ -12,7 +12,7 @@
 import math
 import warnings
 from dataclasses import field
-from typing import Any, Dict, List, Optional, Tuple, Union
+from typing import Any, Dict, List, Optional, Tuple, TYPE_CHECKING, Union
 
 import torch
 import tqdm
@@ -34,7 +34,9 @@
 from pytorch3d.renderer import utils as rend_utils
 
 from pytorch3d.renderer.cameras import CamerasBase
-from visdom import Visdom
+
+if TYPE_CHECKING:
+    from visdom import Visdom
 
 from .base_model import ImplicitronModelBase, ImplicitronRender
 from .feature_extractor import FeatureExtractorBase
@@ -544,7 +546,7 @@ def _get_objective(self, preds) -> Optional[torch.Tensor]:
 
     def visualize(
         self,
-        viz: Visdom,
+        viz: Optional["Visdom"],
         visdom_env_imgs: str,
         preds: Dict[str, Any],
         prefix: str,
@@ -559,7 +561,7 @@ def visualize(
             preds: predictions dict like returned by forward()
             prefix: prepended to the names of images
         """
-        if not viz.check_connection():
+        if viz is None or not viz.check_connection():
             logger.info("no visdom server! -> skipping batch vis")
             return
 

pytorch3d/implicitron/models/visualization/render_flyaround.py

@@ -10,7 +10,7 @@
 import math
 import os
 import random
-from typing import Any, Dict, List, Optional, Sequence, Tuple, Union
+from typing import Any, Dict, List, Optional, Sequence, Tuple, TYPE_CHECKING, Union
 
 import numpy as np
 import torch
@@ -27,7 +27,9 @@
     make_depth_image,
 )
 from tqdm import tqdm
-from visdom import Visdom
+
+if TYPE_CHECKING:
+    from visdom import Visdom
 
 logger = logging.getLogger(__name__)
 
@@ -272,7 +274,7 @@ def _stack_images(ims: torch.Tensor, size: Optional[Tuple[int, int]]) -> torch.T
 def _show_predictions(
     preds: List[Dict[str, Any]],
     sequence_name: str,
-    viz: Visdom,
+    viz: "Visdom",
     viz_env: str = "visualizer",
     predicted_keys: Sequence[str] = (
         "images_render",
@@ -318,7 +320,7 @@ def _show_predictions(
 def _generate_prediction_videos(
     preds: List[Dict[str, Any]],
     sequence_name: str,
-    viz: Optional[Visdom] = None,
+    viz: Optional["Visdom"] = None,
     viz_env: str = "visualizer",
     predicted_keys: Sequence[str] = (
         "images_render",

pytorch3d/implicitron/tools/stats.py

@@ -337,7 +337,7 @@ def plot_stats(
         novisdom = False
 
         viz = get_visdom_connection(server=visdom_server, port=visdom_port)
-        if not viz.check_connection():
+        if viz is None or not viz.check_connection():
             print("no visdom server! -> skipping visdom plots")
             novisdom = True
 

pytorch3d/implicitron/tools/vis_utils.py

@@ -5,10 +5,12 @@
 # LICENSE file in the root directory of this source tree.
 
 import logging
-from typing import Any, Dict, Tuple
+from typing import Any, Dict, Optional, Tuple, TYPE_CHECKING
 
 import torch
-from visdom import Visdom
+
+if TYPE_CHECKING:
+    from visdom import Visdom
 
 
 logger = logging.getLogger(__name__)
@@ -40,9 +42,9 @@ def get_visdom_env(visdom_env: str, exp_dir: str) -> str:
 def get_visdom_connection(
     server: str = "http://localhost",
     port: int = 8097,
-) -> Visdom:
+) -> Optional["Visdom"]:
     """
-    Obtain a connection to a visdom server.
+    Obtain a connection to a visdom server if visdom is installed.
 
     Args:
         server: Server address.
@@ -51,14 +53,23 @@ def get_visdom_connection(
     Returns:
         connection: The connection object.
     """
+    try:
+        from visdom import Visdom
+    except ImportError:
+        logger.debug("Cannot load visdom")
+        return None
+
+    if server == "None":
+        return None
+
     global _viz_singleton
     if _viz_singleton is None:
         _viz_singleton = Visdom(server=server, port=port)
     return _viz_singleton
 
 
 def visualize_basics(
-    viz: Visdom,
+    viz: "Visdom",
     preds: Dict[str, Any],
     visdom_env_imgs: str,
     title: str = "",