Merge branch 'enricoros:main' into main

docs/config-local-localai.md

@@ -1,34 +1,61 @@
-# Local LLM integration with `localai`
+# Run your models with `LocalAI` x `big-AGI`
 
-Integrate local Large Language Models (LLMs) with [LocalAI](https://localai.io).
+[LocalAI](https://localai.io) lets you run your AI models locally, or in the cloud. It supports text, image, asr, speech, and more models.
 
-_Last updated Nov 7, 2023_
+We are deepening the integration between the two products. As of the time of writing, we integrate the following features:
 
-## Instructions
+- ✅ [Text generation](https://localai.io/features/text-generation/) with GPTs
+- ✅ [Function calling](https://localai.io/features/openai-functions/) by GPTs 🆕
+- ✅ [Model Gallery](https://localai.io/models/) to list and install models
+- ✖️ [Vision API](https://localai.io/features/gpt-vision/) for image chats
+- ✖️ [Image generation](https://localai.io/features/image-generation) with stable diffusion
+- ✖️ [Audio to Text](https://localai.io/features/audio-to-text/)
+- ✖️ [Text to Audio](https://localai.io/features/text-to-audio/)
+- ✖️ [Embeddings generation](https://localai.io/features/embeddings/)
+- ✖️ [Constrained grammars](https://localai.io/features/constrained_grammars/) (JSON output)
+- ✖️ Voice cloning 🆕
+
+_Last updated Feb 21, 2024_
+
+## Guide
 
 ### LocalAI installation and configuration
 
 Follow the guide at: https://localai.io/basics/getting_started/
 
-For instance with [Use luna-ai-llama2 with docker compose](https://localai.io/basics/getting_started/#example-use-luna-ai-llama2-model-with-docker-compose):
-
-- clone LocalAI
-- get the model
-- copy the prompt template
-- start docker
-    - -> the server will be listening on `localhost:8080`
-    - verify it works by going to [http://localhost:8080/v1/models](http://localhost:8080/v1/models) on
-      your browser and seeing listed the model you downloaded
+- verify it works by browsing to [http://localhost:8080/v1/models](http://localhost:8080/v1/models)
+  (or the IP:Port of the machine, if running remotely) and seeing listed the model(s) you downloaded
+  listed in the JSON response.
 
-### Integrating LocalAI with big-AGI
+### Integration: chat with LocalAI
 
 - Go to Models > Add a model source of type: **LocalAI**
-- Enter the address: `http://localhost:8080` (default)
-    - If running remotely, replace localhost with the IP of the machine. Make sure to use the **IP:Port** format
-- Load the models
-- Select model & Chat
-
-> NOTE: LocalAI does not list details about the mdoels. Every model is assumed to be
-> capable of chatting, and with a context window of 4096 tokens.
-> Please update the [src/modules/llms/transports/server/openai/models.data.ts](../src/modules/llms/server/openai/models.data.ts)
-> file with the mapping information between LocalAI model IDs and names/descriptions/tokens, etc.
+- Enter the default address: `http://localhost:8080`, or the address of your localAI cloud instance
+  ![configure models](pixels/config-localai-1-models.png)
+  - If running remotely, replace localhost with the IP of the machine. Make sure to use the **IP:Port** format
+- Load the models (click on `Models 🔄`)
+- Select the model and chat
+
+### Integration: Models Gallery
+
+If the running LocalAI instance is configured with a [Model Gallery](https://localai.io/models/):
+
+- Go to Models > LocalAI
+- Click on `Gallery Admin`
+- Select the models to install, and view installation progress
+  ![img.png](pixels/config-localai-2-gallery.png)
+
+## Troubleshooting
+
+##### Unknown Context Window Size
+
+At the time of writing, LocalAI does not publish the model `context window size`.
+Every model is assumed to be capable of chatting, and with a context window of 4096 tokens.
+Please update the [src/modules/llms/transports/server/openai/models.data.ts](../src/modules/llms/server/openai/models.data.ts)
+file with the mapping information between LocalAI model IDs and names/descriptions/tokens, etc.
+
+# 🤝 Support
+
+- Hop into the [LocalAI Discord](https://discord.gg/uJAeKSAGDy) for support and questions
+- Hop into the [big-AGI Discord](https://discord.gg/MkH4qj2Jp9) for questions
+- For big-AGI support, please open an issue in our [big-AGI issue tracker](https://bit.ly/agi-request)

next.config.mjs

@@ -1,18 +1,21 @@
-const BuildOptions = {
-  standalone: !!process.env.BUILD_STANDALONE,
-};
+// Non-default build types
+const buildType =
+  process.env.BIG_AGI_BUILD === 'standalone' ? 'standalone'
+    : process.env.BIG_AGI_BUILD === 'static' ? 'export'
+      : undefined;
+
+buildType && console.log(`   🧠 big-AGI: building for ${buildType}...\n`);
 
 /** @type {import('next').NextConfig} */
 let nextConfig = {
   reactStrictMode: true,
 
-  // [exporting] https://nextjs.org/docs/advanced-features/static-html-export
-  ...BuildOptions.standalone && {
-    // Export the frontend to ./dist
-    output: 'standalone',
+  // [exports] https://nextjs.org/docs/advanced-features/static-html-export
+  ...buildType && {
+    output: buildType,
     distDir: 'dist',
 
-    // Disable Image optimization
+    // disable image optimization for exports
     images: { unoptimized: true },
 
     // Optional: Change links `/me` -> `/me/` and emit `/me.html` -> `/me/index.html`
@@ -34,8 +37,9 @@ let nextConfig = {
       layers: true,
     };
 
-    // [exporting] prevent too many small files (50kb)
-    BuildOptions.standalone && (config.optimization.splitChunks.minSize = 50 * 1024);
+    // prevent too many small chunks (40kb min) on 'client' packs (not 'server' or 'edge-server')
+    if (typeof config.optimization.splitChunks === 'object' && config.optimization.splitChunks.minSize)
+      config.optimization.splitChunks.minSize = 40 * 1024;
 
     return config;
   },