From 4b338a3abd9f95c1bf186c9954ba7a1b2a66fde6 Mon Sep 17 00:00:00 2001
From: Evening <johnchangqi@hotmail.com>
Date: Sun, 26 Feb 2023 15:40:42 +0800
Subject: [PATCH] Update README.md

---
 README.md | 45 +++++++++++++++++++++++++++++++++++----------
 1 file changed, 35 insertions(+), 10 deletions(-)

diff --git a/README.md b/README.md
index 6140d306..f96aecc9 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
-# [**Try Out Opal**](https://colab.research.google.com/drive/17WRYX24NJ4JaecJh8rUQ0iXSO93LrBUY?usp=sharing)
+# :arrow_forward: [**Try Out Opal on Streamlit**](https://opal-ai.streamlit.app/)
 
-# opal
+# :comet: opal
 opal is an accuracy-prediction model.
 
 It uses Neural Collaborative Filtering to learn associations between user and maps, then using those associations to
@@ -9,13 +9,13 @@ predict new scores never before seen.
 **Performance Error Graph**
 ![Performance Graph](models/V2_2023_01/error.png)
 
-## Project Status
+## :hourglass_flowing_sand: Project Status
 
 Currently, it's in its early access, that means, it'll have many problems!
 
 However, we're working on it to minimize these issues o wo)b
 
-## Dataset Used
+## :arrow_double_down: Dataset Used
 
 I used the top 10K mania users data from https://data.ppy.sh.
 After preprocessing, we use
@@ -29,15 +29,40 @@ This models can thus help predict >300m unplayed scores!
 We deem a player on separate years as a different user. This is to reflect
 the improvement of the player after time.
 
-## Usage
+## :high_brightness: Usage
 
-Currently, I'm only providing the solution as a **PyTorch Lightning Checkpoint**. 
-That is, you **need** to use **Python** to use this!
+If you want to use this locally, you need Python and the packages listed in [requirements.txt](requirements.txt)
 
-I'm working on making the model more accurate before making the model UI through Streamlit.
-So hang on tight!
+Below is a recipe on how to use it.
+```py
+path_to_model = MODEL_DIR / "V2_2023_01/checkpoints/epoch=5-step=43584.ckpt"
+net = NeuMF.load_from_checkpoint(path_to_model.as_posix())
 
-## Why not Score Metric?
+# THIS MUST BE RAN TO AVOID TRAINING THE MODEL
+net.eval()
+
+pred = net.predict('<USER_ID>/<YEAR>', '<MAP_ID>/<SPEED>')
+preds = net.predict(['<USER_ID>/<YEAR>', '<USER_ID>/<YEAR>', ...],
+                    ['<MAP_ID>/<SPEED>', '<MAP_ID>/<SPEED>', ...])
+# <SPEED> must be -1, 0, or 1. Where -1 is half time, 0 is normal time, 1 is double time.
+
+# E.g.
+# Predict Evening on Year 2017, on the map Triumph & Regret [Regret] at Double Time
+pred = net.predict('2193881/2017', '767046/1')
+
+# Note that you can, and should predict in a list format, so it's a significantly faster.
+# Note that the map and user id can be different!
+preds = net.predict(['2193881/2017', '2193881/2018'], ['767046/1', '767046/0'])
+```
+
+## :brain: AlphaOsu!
+Currently, opal doesn't provide recommendations, however, you can try out [AlphaOsu!](https://alphaosu.keytoix.vip/).
+- [AlphaOsu! GitHub](https://github.com/AlphaOSU)
+- [Support AlphaOsu!](https://alphaosu.keytoix.vip/support)
+
+## Annex
+
+### Why not Score Metric?
 Score is not straightforward to calculate, and may be difficult to debug. Furthermore, score isn't of interest when
 calculating performance points anymore.