update readme

README.md

@@ -1 +1,13 @@
 # fsrs-browser
+
+This project runs [fsrs-rs](https://github.com/open-spaced-repetition/fsrs-rs) in the browser with support for training FSRS parameters.
+
+It git submodules [fsrs-rs](https://github.com/AlexErrant/fsrs-rs/tree/fsrs-browser) and [burn](https://github.com/AlexErrant/burn/tree/fsrs-browser) for reasons given [here](https://github.com/Tracel-AI/burn/pull/938#issuecomment-1925913866). As such, it will not be up to date with the latest fsrs-rs.
+
+## Building and demoing
+
+Run `./dev.sh` for fast builds or `./prod.sh` for fast runs.
+
+Run the `/sandbox` project to demo various behavior.
+
+I highly encourage `./prod.sh` if you intend to run training. On my machine training 24,394 FSRS items/revlogs on `./dev` takes days, while `./prod.sh` takes 3.5 seconds.

sandbox/README.md

@@ -1,15 +1,5 @@
 ## Usage
 
-Those templates dependencies are maintained via [pnpm](https://pnpm.io) via `pnpm up -Lri`.
-
-This is the reason you see a `pnpm-lock.yaml`. That being said, any package manager will work. This file can be safely be removed once you clone a template.
-
-```bash
-$ npm install # or pnpm install or yarn install
-```
-
-### Learn more on the [Solid Website](https://solidjs.com) and come chat with us on our [Discord](https://discord.com/invite/solidjs)
-
 ## Available Scripts
 
 In the project directory, you can run:
@@ -28,7 +18,3 @@ It correctly bundles Solid in production mode and optimizes the build for the be
 
 The build is minified and the filenames include the hashes.<br>
 Your app is ready to be deployed!
-
-## Deployment
-
-You can deploy the `dist` folder to any static host provider (netlify, surge, now, etc.)

sandbox/src/train.ts

@@ -23,6 +23,7 @@ async function loadSqliteAndRun(ab: ArrayBuffer) {
 	await wasm()
 	await initThreadPool(navigator.hardwareConcurrency)
 	await sleep(1000) // the workers need time to spin up. TODO, post an init message and await a response. Also maybe move worker construction to Javascript.
+	console.time('full training time')
 	let fsrs = new Fsrs()
 	const db = await getDb(ab)
 	let baseQuery = `FROM revlog
@@ -77,7 +78,9 @@ async function loadSqliteAndRun(ab: ArrayBuffer) {
 			types,
 			usns,
 		)
+		console.timeEnd('full training time')
 		console.log('trained weights are', weights)
+		console.log('revlog count', count)
 	} finally {
 		db.close()
 	}