feat: use cutom WebLLM build to have qwen-2-0.5b quantized versions

Author: Neet-Nestor

app/client/api.ts

@@ -1,4 +1,7 @@
-import { ChatCompletionFinishReason, CompletionUsage } from "@mlc-ai/web-llm";
+import {
+  ChatCompletionFinishReason,
+  CompletionUsage,
+} from "@neet-nestor/web-llm";
 import { CacheType, Model } from "../store";
 export const ROLES = ["system", "user", "assistant"] as const;
 export type MessageRole = (typeof ROLES)[number];

app/client/mlcllm.ts

@@ -1,6 +1,9 @@
 import log from "loglevel";
 import { ChatOptions, LLMApi } from "./api";
-import { ChatCompletionFinishReason, CompletionUsage } from "@mlc-ai/web-llm";
+import {
+  ChatCompletionFinishReason,
+  CompletionUsage,
+} from "@neet-nestor/web-llm";
 
 export class MlcLLMApi implements LLMApi {
   private endpoint: string;

app/client/webllm.ts

@@ -12,10 +12,10 @@ import {
   WebWorkerMLCEngine,
   CompletionUsage,
   ChatCompletionFinishReason,
-} from "@mlc-ai/web-llm";
+} from "@neet-nestor/web-llm";
 
 import { ChatOptions, LLMApi, LLMConfig, RequestMessage } from "./api";
-import { LogLevel } from "@mlc-ai/web-llm";
+import { LogLevel } from "@neet-nestor/web-llm";
 import { fixMessage } from "../utils";
 import { DEFAULT_MODELS } from "../constant";
 

app/components/home.tsx

@@ -13,7 +13,7 @@ import {
   Route,
   useLocation,
 } from "react-router-dom";
-import { ServiceWorkerMLCEngine } from "@mlc-ai/web-llm";
+import { ServiceWorkerMLCEngine } from "@neet-nestor/web-llm";
 
 import MlcIcon from "../icons/mlc.svg";
 import LoadingIcon from "../icons/three-dots.svg";

app/components/settings.tsx

@@ -34,7 +34,7 @@ import { ErrorBoundary } from "./error";
 import { InputRange } from "./input-range";
 import { useNavigate } from "react-router-dom";
 import { nanoid } from "nanoid";
-import { LogLevel } from "@mlc-ai/web-llm";
+import { LogLevel } from "@neet-nestor/web-llm";
 import { WebLLMContext } from "../context";
 
 function EditPromptModal(props: { id: string; onClose: () => void }) {

app/store/chat.ts

@@ -16,7 +16,10 @@ import { estimateTokenLength } from "../utils/token";
 import { nanoid } from "nanoid";
 import { createPersistStore } from "../utils/store";
 import { WebLLMApi } from "../client/webllm";
-import { ChatCompletionFinishReason, CompletionUsage } from "@mlc-ai/web-llm";
+import {
+  ChatCompletionFinishReason,
+  CompletionUsage,
+} from "@neet-nestor/web-llm";
 
 export type ChatMessage = RequestMessage & {
   date: string;

app/store/config.ts

@@ -1,4 +1,4 @@
-import { LogLevel } from "@mlc-ai/web-llm";
+import { LogLevel } from "@neet-nestor/web-llm";
 import { ModelRecord } from "../client/api";
 import {
   DEFAULT_INPUT_TEMPLATE,

app/worker/service-worker.ts

@@ -1,4 +1,4 @@
-import { ServiceWorkerMLCEngineHandler } from "@mlc-ai/web-llm";
+import { ServiceWorkerMLCEngineHandler } from "@neet-nestor/web-llm";
 import { defaultCache } from "@serwist/next/worker";
 import type { PrecacheEntry, SerwistGlobalConfig } from "serwist";
 import { CacheFirst, ExpirationPlugin, Serwist } from "serwist";

app/worker/web-worker.ts

@@ -1,5 +1,5 @@
 import log from "loglevel";
-import { WebWorkerMLCEngineHandler } from "@mlc-ai/web-llm";
+import { WebWorkerMLCEngineHandler } from "@neet-nestor/web-llm";
 
 let handler: WebWorkerMLCEngineHandler;
 

package.json

@@ -17,7 +17,7 @@
   "dependencies": {
     "@fortaine/fetch-event-source": "^3.0.6",
     "@hello-pangea/dnd": "^16.5.0",
-    "@mlc-ai/web-llm": "^0.2.46",
+    "@neet-nestor/web-llm": "^0.2.58",
     "@serwist/next": "^9.0.2",
     "@svgr/webpack": "^6.5.1",
     "emoji-picker-react": "^4.9.2",