Document `--use-cog-base-image`

[zeke]

There's a new flag for cog push that lets you use prebuilt base images designed specifically for Cog.

Using this flag when pushing your models can significantly reduce their cold boot time:

cog push --use-cog-base-image r8.im/my-name/my-model

For more details, see #1778

Tasks

• Add a new doc that explains what prebuilt Cog images are, what's on them, etc.
• Clarify relationship between use-cog-base-image and use-cuda-base-image. Are they related? Do they override each other?

  $ cog push --help | grep base-image
    --use-cog-base-image           Use pre-built Cog base image for faster cold boots (default true)
    --use-cuda-base-image string   Use Nvidia CUDA base image, 'true' (default) or 'false' (use python base image). False results in a smaller image but may cause problems for non-torch projects (default "auto")
    ```
  

• Update getting started docs with correct cog push incantations.