Continue brain.js refactoring, but it's still not finished yet.

server/settings.default.toml

@@ -65,8 +65,14 @@ network_arch = [ 64 ]
 
 # The number of epochs to train for.
 epochs = 1000
-# The percentage, between 0 and 1, of data that should be set aside for validation.
-validation_split = 0.1
+# Cut training short if the mean squared error drops below 
+# this value
+error_threshold = 0.0001
+
+# The learning rate that the neural networks should learn at
+learning_rate = 0.1
+# The momentum term to use when learning
+momentum = 0.1
 
 # The directory to output trained AIs to, relative to the repository root.
 output_directory = "app/ais/"

server/train-ai/AITrainer.mjs

@@ -70,7 +70,7 @@ class AITrainer {
 	 */
 	async train_gateway(gateway_id, destination_filename) {
 		this.l.log(`${this.a.fgreen}${this.a.hicol}Training AI for gateway ${gateway_id}${this.a.reset}`);
-		let model = this.generate_neural_net();
+		let net = this.generate_neural_net();
 		
 		// let dataset_input = tf.data.generator(
 		// 	this.dataset_fetcher.fetch_input.bind(this.dataset_fetcher, gateway_id)
@@ -85,17 +85,17 @@ class AITrainer {
 		// 	}).shuffle(this.settings.ai.batch_size * 4)
 		// 	.batch(this.settings.ai.batch_size);
 		// 
-		let datasets = this.dataset_fetcher.fetch_all(gateway_id);
+		let dataset = this.dataset_fetcher.fetch_all(gateway_id);
 		
-		let result = await model.fit(
-			tf.tensor(datasets.input),
-			tf.tensor(datasets.output),
-			{
-				epochs: this.settings.ai.epochs,
-				batchSize:  this.settings.ai.batch_size,
-				validationSplit: this.settings.ai.validation_split
-			}
-		);
+		let result = net.train(dataset, {
+			iterations: this.settings.ai.epochs,
+			errorThresh: this.settings.ai.error_threshold,
+			
+			learningRate: this.settings.ai.learning_rate,
+			momentum: this.settings.ai.momentum,
+			
+			timeout: Infinity
+		});
 		
 		await model.save(`file://${destination_filename}`);
 		// console.log(result);

server/train-ai/DatasetFetcher.mjs

@@ -28,10 +28,7 @@ class DatasetFetcher {
 	fetch_all(gateway_id) {
 		let gateway_location = this.repo_gateway.get_by_id(gateway_id);
 		
-		let result = {
-			input: [],
-			output: []
-		};
+		let result = [];
 		for(let rssi of this.repo_rssi.iterate_gateway(gateway_id)) {
 			let next_input = this.normalise_latlng(rssi.latitude, rssi.longitude);
 			let distance_from_gateway = haversine(gateway_location, rssi);
@@ -47,12 +44,15 @@ class DatasetFetcher {
 			
 			console.log(`Distance from gateway: ${haversine(gateway_location, rssi)}m`);
 			
-			result.output.push([
-				clamp(normalise(rssi.rssi,
-					{ min: this.settings.ai.rssi_min, max: this.settings.ai.rssi_max },
-					{ min: 0, max: 1 }
-				), 0, 1)
-			]);
+			let next_output = clamp(normalise(rssi.rssi,
+				{ min: this.settings.ai.rssi_min, max: this.settings.ai.rssi_max },
+				{ min: 0, max: 1 }
+			), 0, 1);
+			
+			result.push({
+				input: next_input,
+				output: next_output
+			});
 		}
 		
 		for(let reading of this.repo_reading.iterate_unreceived()) {