Update index.js

index.js

@@ -1,4 +1,4 @@
-import { env, AutoTokenizer, RawImage, Tensor } from 'https://cdn.jsdelivr.net/npm/@huggingface/transformers';
+import { env, AutoTokenizer, RawImage, Tensor, getSession } from 'https://cdn.jsdelivr.net/npm/@huggingface/transformers';
 import { getModelJSON } from "https://cdn.jsdelivr.net/npm/@huggingface/transformers@3.0.2/src/utils/hub.js";
 import * as ort from "https://cdn.jsdelivr.net/npm/onnxruntime-web@1.20.0/dist/ort.webgpu.mjs";
 
@@ -22,6 +22,9 @@ const BASE_MODEL = "Qwen/Qwen2-VL-2B-Instruct";
 const QUANTIZATION = "q4f16";
 const MAX_SINGLE_CHAT_LENGTH = 10;
 
+// const ONNX_MODEL_BASE_URL= 
+console.log(getSession);
+
 status.textContent = 'Loading model...';
 status.textContent = 'Ready';
 
@@ -124,7 +127,7 @@ export async function imageTextToText(
     const pixel_values = image.unsqueeze(0);
 
     const ortSessionA = await ort.InferenceSession.create(
-      `${BASE_URL}/QwenVL_A${suffix}.onnx`,
+      `${ONNX_MODEL_BASE_URL}/QwenVL_A${suffix}.onnx`,
       { executionProviders: ["webgpu"] }
     );