merge: branch 'main' into lab/pred

ml/pred/dataset.py

@@ -20,7 +20,7 @@ class VideoPlayDataset(Dataset):
         self.valid_series = [s for s in self.series_dict.values() if len(s['abs_time']) > 1]
         self.term = term
         # Set time window based on term
-        self.time_window = 1000 * 24 * 3600 if term == 'long' else 7 * 24 * 3600
+        self.time_window = 1000 * 24 * 3600 if term == 'long' else 3 * 24 * 3600
         MINUTE = 60
         HOUR = 3600
         DAY = 24 * HOUR
@@ -37,6 +37,7 @@ class VideoPlayDataset(Dataset):
             ]
         else:
             self.feature_windows = [
+                #( 5 * MINUTE,  0 * MINUTE),
                 ( 15 * MINUTE,  0 * MINUTE),
                 ( 40 * MINUTE,  0 * MINUTE),
                 ( 1 * HOUR,  0 * HOUR),
@@ -45,7 +46,7 @@ class VideoPlayDataset(Dataset):
                 ( 3 * HOUR,  0 * HOUR),
                 #( 6 * HOUR,  3 * HOUR),
                 ( 6 * HOUR,  0 * HOUR),
-                (18 * HOUR, 12 * HOUR),
+                #(18 * HOUR, 12 * HOUR),
                 #( 1 * DAY,   6 * HOUR),
                 ( 1 * DAY,   0 * DAY),
                 #( 2 * DAY,   1 * DAY),

ml/pred/inference.py

@@ -4,20 +4,20 @@ from model import CompactPredictor
 import torch
 
 def main():
-    model = CompactPredictor(10).to('cpu', dtype=torch.float32)
-    model.load_state_dict(torch.load('./pred/checkpoints/long_term.pt'))
+    model = CompactPredictor(15).to('cpu', dtype=torch.float32)
+    model.load_state_dict(torch.load('./pred/checkpoints/model_20250320_0045.pt'))
     model.eval()
     # inference
-    initial = 997029
+    initial = 999704
     last = initial
-    start_time = '2025-03-17 00:13:17'
-    for i in range(1, 120):
-        hour = i / 0.5
+    start_time = '2025-03-19 22:00:42'
+    for i in range(1, 48):
+        hour = i / 6
         sec = hour * 3600
         time_d = np.log2(sec)
         data = [time_d, np.log2(initial+1), # time_delta, current_views
-                6.111542, 8.404707, 10.071566, 11.55888, 12.457823,# grows_feat
-                 0.009225, 0.001318, 28.001814# time_feat
+                4.857981, 6.29067, 6.869476, 6.58392, 6.523051, 8.242355, 8.841574, 10.203909, 11.449314, 12.659556, # grows_feat
+                0.916956, 0.416708, 28.003162 # time_feat
         ]
         np_arr = np.array([data])
         tensor = torch.from_numpy(np_arr).to('cpu', dtype=torch.float32)
@@ -25,7 +25,7 @@ def main():
         num = output.detach().numpy()[0][0]
         views_pred = int(np.exp2(num)) + initial
         current_time = datetime.datetime.strptime(start_time, '%Y-%m-%d %H:%M:%S') + datetime.timedelta(hours=hour)
-        print(current_time.strftime('%m-%d %H:%M:%S'), views_pred, views_pred - last)
+        print(current_time.strftime('%m-%d %H:%M'), views_pred, views_pred - last)
         last = views_pred
 
 if __name__ == '__main__':

ml/pred/train.py

@@ -38,7 +38,7 @@ def train(model, dataloader, device, epochs=100):
     scheduler = torch.optim.lr_scheduler.OneCycleLR(optimizer, max_lr=1e-3,
                                                   total_steps=len(dataloader)*30)
     # Huber loss
-    criterion = asymmetricHuberLoss(delta=1.0, beta=2.1)
+    criterion = asymmetricHuberLoss(delta=1.0, beta=2.2)
     
     model.train()
     global_step = 0
@@ -100,7 +100,7 @@ if __name__ == "__main__":
     device = 'mps'
     
     # Initialize dataset and model
-    dataset = VideoPlayDataset('./data/pred', './data/pred/publish_time.csv', 'short')
+    dataset = VideoPlayDataset('./data/pred', './data/pred/publish_time.csv', 'short', 712)
     dataloader = DataLoader(dataset, batch_size=128, shuffle=True, collate_fn=collate_fn)
     
     # Get feature dimension

packages/backend/database.ts

@@ -1,10 +1,12 @@
 import { type Client, Pool } from "https://deno.land/x/postgres@v0.19.3/mod.ts";
-import { postgresConfig } from "@core/db/pgConfig.ts";
+import { postgresConfig, postgresConfigCred } from "@core/db/pgConfig.ts";
 import { createMiddleware } from "hono/factory";
 
 const pool = new Pool(postgresConfig, 4);
+const poolCred = new Pool(postgresConfigCred, 2);
 
 export const db = pool;
+export const dbCred = poolCred;
 
 export const dbMiddleware = createMiddleware(async (c, next) => {
     const connection = await pool.connect();
@@ -13,8 +15,16 @@ export const dbMiddleware = createMiddleware(async (c, next) => {
 	connection.release();
 });
 
+export const dbCredMiddleware = createMiddleware(async (c, next) => {
+    const connection = await poolCred.connect();
+	c.set("dbCred", connection);
+	await next();
+	connection.release();
+})
+
 declare module "hono" {
 	interface ContextVariableMap {
 		db: Client;
+		dbCred: Client;
 	}
 }

packages/backend/deno.json

@@ -1,6 +1,7 @@
 {
 	"name": "@cvsa/backend",
 	"imports": {
+		"@rabbit-company/argon2id": "jsr:@rabbit-company/argon2id@^2.1.0",
 		"hono": "jsr:@hono/hono@^4.7.5",
 		"zod": "npm:zod",
 		"yup": "npm:yup"

packages/backend/main.ts

@@ -1,15 +1,18 @@
 import { Hono } from "hono";
-import { dbMiddleware } from "./database.ts";
+import { dbCredMiddleware, dbMiddleware } from "./database.ts";
 import { rootHandler } from "./root.ts";
 import { getSnapshotsHanlder } from "./snapshots.ts";
+import { registerHandler } from "./register.ts";
 
 export const app = new Hono();
 
 app.use('/video/*', dbMiddleware);
+app.use('/user', dbCredMiddleware);
 
 app.get("/", ...rootHandler);
 
 app.get('/video/:id/snapshots', ...getSnapshotsHanlder);
+app.post('/user', ...registerHandler);
 
 const fetch = app.fetch;
 
@@ -17,4 +20,4 @@ export default {
 	fetch,
 } satisfies Deno.ServeDefaultExport;
 
-export const VERSION = "0.2.4";
+export const VERSION = "0.3.0";

packages/backend/register.ts

@@ -0,0 +1,65 @@
+import { createHandlers } from "./utils.ts";
+import Argon2id from "@rabbit-company/argon2id";
+import { object, string, ValidationError } from "yup";
+import type { Context } from "hono";
+import type { Bindings, BlankEnv, BlankInput } from "hono/types";
+import type { Client } from "https://deno.land/x/postgres@v0.19.3/mod.ts";
+
+const RegistrationBodySchema = object({
+	username: string().trim().required("Username is required").max(50, "Username cannot exceed 50 characters"),
+	password: string().required("Password is required"),
+    nickname: string().optional(),
+});
+
+type ContextType = Context<BlankEnv & { Bindings: Bindings }, "/user", BlankInput>;
+
+export const userExists = async (username: string, client: Client) => {
+	const query = `
+        SELECT * FROM users WHERE username = $1
+    `;
+	const result = await client.queryObject(query, [username]);
+	return result.rows.length > 0;
+}
+
+export const registerHandler = createHandlers(async (c: ContextType) => {
+	const client = c.get("dbCred");
+
+	try {
+		const body = await RegistrationBodySchema.validate(await c.req.json());
+		const { username, password, nickname } = body;
+
+        if (await userExists(username, client)) {
+			return c.json({
+				message: `User "${username}" already exists.`,
+			}, 400);
+        }
+
+		const hash = await Argon2id.hashEncoded(password);
+
+		const query = `
+            INSERT INTO users (username, password, nickname) VALUES ($1, $2, $3)
+        `;
+		await client.queryObject(query, [username, hash, nickname || null]);
+
+		return c.json({
+			message: `User "${username}" registered successfully.`,
+		}, 201);
+	} catch (e) {
+		if (e instanceof ValidationError) {
+			return c.json({
+				message: "Invalid registration data.",
+				errors: e.errors,
+			}, 400); 
+		} else if (e instanceof SyntaxError) {
+			return c.json({
+				message: "Invalid JSON in request body.",
+			}, 400);
+		} else {
+			console.error("Registration error:", e);
+			return c.json({
+				message: "An unexpected error occurred during registration.",
+				error: (e as Error).message,
+			}, 500);
+		}
+	}
+});