File size: 4,523 Bytes
1d3f990 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 | /**
* Ahmad Bot Web Worker
* Offloads model inference to prevent UI blocking
* Communicates with main thread via message passing
*/
let engine = null;
let isInitializing = false;
/**
* Handle messages from main thread
*/
self.onmessage = async (event) => {
const { type, data } = event.data;
try {
switch (type) {
case 'INIT':
await handleInit(data);
break;
case 'GENERATE':
await handleGenerate(data);
break;
case 'STOP':
handleStop();
break;
case 'STATUS':
handleStatus();
break;
default:
console.warn(`[Ahmad Bot Worker] Unknown message type: ${type}`);
}
} catch (error) {
self.postMessage({
type: 'ERROR',
error: error.message,
stack: error.stack,
});
}
};
/**
* Initialize WebLLM engine
*/
async function handleInit(data) {
if (isInitializing || engine) {
self.postMessage({
type: 'INIT_ACK',
status: 'ALREADY_INITIALIZED',
});
return;
}
isInitializing = true;
self.postMessage({
type: 'INIT_PROGRESS',
message: 'Initializing WebLLM engine...',
});
try {
// Import WebLLM if available
const { modelId = 'Llama-2-7b-chat-hf-q4f32_1-MLC' } = data;
// Check if WebLLM is available in worker
if (typeof self.webllm === 'undefined') {
throw new Error('WebLLM not available in worker. Ensure script is imported.');
}
// Initialize engine
engine = new self.webllm.MLCEngine({
model: modelId,
temperature: 0.3,
top_p: 0.85,
max_gen_len: 512,
});
// Wait for ready
await engine.ready;
isInitializing = false;
self.postMessage({
type: 'INIT_ACK',
status: 'READY',
model: modelId,
});
} catch (error) {
isInitializing = false;
throw error;
}
}
/**
* Generate response with streaming
*/
async function handleGenerate(data) {
if (!engine) {
throw new Error('Engine not initialized');
}
const { messages, requestId } = data;
try {
self.postMessage({
type: 'GENERATION_START',
requestId: requestId,
});
let fullResponse = '';
let tokenCount = 0;
// Create stream
const stream = await engine.chat.completions.create({
messages: messages,
stream: true,
temperature: 0.3,
top_p: 0.85,
});
// Consume stream and send tokens
for await (const chunk of stream) {
const token = chunk.choices?.[0]?.delta?.content || '';
if (token) {
fullResponse += token;
tokenCount++;
// Send token to main thread
self.postMessage({
type: 'TOKEN',
requestId: requestId,
token: token,
tokenCount: tokenCount,
});
// Yield to prevent blocking
if (tokenCount % 10 === 0) {
await new Promise(r => setTimeout(r, 0));
}
}
}
// Send completion
self.postMessage({
type: 'GENERATION_COMPLETE',
requestId: requestId,
fullResponse: fullResponse,
tokenCount: tokenCount,
});
} catch (error) {
self.postMessage({
type: 'GENERATION_ERROR',
requestId: data.requestId,
error: error.message,
});
}
}
/**
* Stop current generation
*/
function handleStop() {
if (engine) {
engine.interruptGenerate?.();
}
self.postMessage({
type: 'GENERATION_STOPPED',
});
}
/**
* Get engine status
*/
function handleStatus() {
self.postMessage({
type: 'STATUS_ACK',
isReady: engine !== null,
isInitializing: isInitializing,
});
}
// Signal to main thread that worker is ready
self.postMessage({
type: 'WORKER_READY',
});
|