Update cli docs (#674)

docs/reference/cli/index.md

@@ -0,0 +1,25 @@
+# CLI
+
+We provide a command line interface for training your own NeRFs (no coding necessary). You can learn more about each command by using the `--help` argument.
+
+## Commands
+
+| Command          | Description                                      |
+| ---------------- | ------------------------------------------------ |
+| ns_process_data  | Generate a dataset from your own photos or video |
+| ns_download_data | Download existing captures                       |
+| ns_train         | Generate a NeRF                                  |
+| ns_render        | Render out a video of your NeRF                  |
+| ns_eval          | Run evaluation metrics for your Model            |
+| ns_install_cli   | Install tab completion for all scripts           |
+
+```{toctree}
+:maxdepth: 1
+:hidden:
+
+ns_process_data
+ns_download_data
+ns_train
+ns_render
+ns_eval
+```

docs/reference/cli/index.rst

@@ -1,12 +0,0 @@
-.. _cli:
-
-CLI
-============
-
-.. toctree::
-    :maxdepth: 1
-
-    ns_process_data
-    ns_download_data
-    ns_render
-    ns_eval

docs/reference/cli/ns_download_data.md

@@ -4,6 +4,6 @@
 .. argparse::
     :module: scripts.downloads.download_data
     :func: get_parser_fn
-    :prog: ns-process-data
+    :prog: ns-download-data
     :nodefault:
 ```

docs/reference/cli/ns_eval.md

@@ -4,6 +4,6 @@
 .. argparse::
     :module: scripts.eval
     :func: get_parser_fn
-    :prog: ns-process-data
+    :prog: ns-eval
     :nodefault:
 ```

docs/reference/cli/ns_render.md

@@ -9,6 +9,6 @@ Make sure to have [FFmpeg](https://ffmpeg.org/download.html) installed.
 .. argparse::
     :module: scripts.render
     :func: get_parser_fn
-    :prog: ns-process-data
+    :prog: ns-render
     :nodefault:
 ```

docs/reference/cli/ns_train.md

@@ -0,0 +1,31 @@
+# ns-train
+
+Primary interface for training a NeRF model. `--help` is your friend when navigating command arguments. We also recommend installing the tab completion `ns-install-cli`.
+
+```bash
+usage: ns-train {method} [method args] {dataparser} [dataparser args]
+```
+
+If you are using a nerfstudio data set, the minimal command is:
+
+```bash
+ns-train nerfacto --data YOUR_DATA
+```
+
+To learn about the available methods:
+
+```bash
+ns-train --help
+```
+
+To learn about a methods parameters:
+
+```bash
+ns-train {method} --help
+```
+
+By default the nerfstudio dataparser is used. If you would like to use a different dataparser it can be specified after all of the method arguments. For a list of dataparser options:
+
+```bash
+ns-train {method} {dataparser} --help
+```

scripts/downloads/download_data.py

@@ -55,7 +55,6 @@ def grab_file_id(zip_url: str) -> str:
     return s.split("/")[0]
 
 
-# https://drive.google.com/drive/folders/1Wh66z3qQTZ8o2MwPPwYOrdwQtXUEQFyq?usp=sharing
 nerfstudio_file_ids = {
     "bunny": grab_file_id("https://drive.google.com/file/d/1oVytSgd1gQclF0CybJ6JdXDmaqKOx7Wg/view?usp=sharing"),
     "bww_entrance": grab_file_id("https://drive.google.com/file/d/1ylkRHtfB3n3IRLf2wplpfxzPTq7nES9I/view?usp=sharing"),
@@ -105,7 +104,14 @@ def main(
     capture_name: Optional[DatasetName] = None,  # type: ignore
     save_dir: Path = Path("data/"),
 ):
-    """Main download script to download all data.
+    """Script to download existing datasets.
+
+    We currently support the following datasets:
+
+    - nerfstudio: Growing collection of real-world scenes. Use the `capture_name` argument to specify
+        which capture to download.
+    - blender: Blender synthetic scenes realeased with NeRF.
+    - friends: Friends TV show scenes.
 
     Args:
         dataset: The dataset to download (from).