implement Groq TPM rate limit handling - wait 60s for token budget reset

src/chat.js

@@ -115,17 +115,50 @@ export async function* sendMessage(message) {
 	
 	console.log("Chat using tokens:", finalAvailable, "calculated:", calculatedAvailable, "max completion:", maxCompletionTokens);
 	
-	const response = await fetch(apiUrl, {
-		method: "POST",
-		headers: headers,
-		body: JSON.stringify({
-			model: model,
-			messages: chatHistory,
-			temperature: 0.5,
-			max_tokens: finalAvailable,
-			stream: true,
-		}),
-	});
+	let response;
+	let retryCount = 0;
+	const maxRetries = 3;
+	
+	while (retryCount < maxRetries) {
+		try {
+			response = await fetch(apiUrl, {
+				method: "POST",
+				headers: headers,
+				body: JSON.stringify({
+					model: model,
+					messages: chatHistory,
+					temperature: 0.5,
+					max_tokens: finalAvailable,
+					stream: true,
+				}),
+			});
+			
+			// Check for rate limit error
+			if (response.status === 429 || response.status === 400) {
+				const errorText = await response.text();
+				if (errorText.includes('rate_limit_exceeded')) {
+					console.log(`Rate limit hit, waiting 60 seconds before retry ${retryCount + 1}/${maxRetries}...`);
+					yield "⏳ Rate limit reached, waiting 60 seconds for TPM window to reset...";
+					await new Promise(resolve => setTimeout(resolve, 60000)); // Wait 60 seconds for TPM reset
+					retryCount++;
+					continue;
+				} else {
+					// Non-rate-limit error, throw it
+					throw new Error(`API Error ${response.status}: ${errorText}`);
+				}
+			}
+			
+			// If we get here, either success or non-rate-limit error
+			break;
+		} catch (error) {
+			console.error("Fetch error:", error);
+			throw error;
+		}
+	}
+	
+	if (retryCount >= maxRetries) {
+		throw new Error("Rate limit exceeded after multiple retries. Please try again later.");
+	}
 
 	const reader = response.body.getReader();
 	const decoder = new TextDecoder();