Merge pull request #23 from wandb/fix/weave-logging

Fix ollama, better logging, improved README

.devcontainer/devcontainer.json

@@ -45,7 +45,10 @@
     }
   },
 
+  // Install Ollama in the prebuild step
+  "onCreateCommand": "curl -fsSL https://ollama.com/install.sh | sh",
   "postCreateCommand": "bash ./.devcontainer/postCreateCommand.sh",
+  "postStartCommand": "nohup bash -c 'ollama serve &'",
 
   "secrets": {
     "OPENAI_API_KEY": {

.devcontainer/postCreateCommand.sh

@@ -4,9 +4,6 @@
 mkdir -p $HOME/.cache
 sudo chown -R $USER $HOME/.cache
 
-# Ollama time
-curl -fsSL https://ollama.com/install.sh | sh
-
 # Install node packages
 cd /workspaces/openui/frontend
 pnpm install
@@ -15,6 +12,9 @@ pnpm install
 cd /workspaces/openui/backend
 pip install -e .[test]
 
+# Pull a model for ollama, using llava for now as it's multi-modal
+ollama pull llava
+
 # addressing warning...
 git config --unset-all core.hooksPath
 pre-commit install --allow-missing-config

README.md

@@ -56,6 +56,10 @@ npm run dev
 
 This should open another service on port 5173, that's the service you'll want to visit.  All changes to both the frontend and backend will automatically be reloaded and reflected in your browser.
 
+### Ollama
+
+The codespace should have ollama running and automatically downloads the `llava` model.  You can select this model from the settings gear icon in the upper left corner of the application.  Any models you pull i.e. `ollama pull llama` will show up in the settings modal.
+
 ### Resources
 
 See the readmes in the [frontend](./frontend/README.md) and [backend](./backend/README.md) directories.