decidekit 0.2.1 → 0.2.2

package/CHANGELOG.md

@@ -1,5 +1,12 @@
 # Changelog
 
+## 0.2.2 (2026-11-02)
+
+### Browser
+
+- New option `wasmHost` serves onnxruntime's wasm from your own host instead of jsDelivr. It applies to the whole page or process.
+- The docs now say the wasm comes from jsDelivr. The FAQ on network requests left that out before.
+
 ## 0.2.1 (2026-11-02)
 
 ### Browser

package/README.md

@@ -155,7 +155,7 @@ decidekit runs the Decide models, each a sentence encoder plus a calibrated head
 | Decide Multilingual | [`debuffjp/decide-multilingual`](https://huggingface.co/debuffjp/decide-multilingual) | about 118 MB | 76% on Japanese and Chinese intents | intfloat's [multilingual-e5-small](https://huggingface.co/intfloat/multilingual-e5-small), identical to `Xenova/multilingual-e5-small` |
 
 Pass the id as `model` to `createDecider()`. The files download from Hugging Face on first use, pinned to a revision, and
-are then cached. To serve them yourself, pass `modelHost`: see [self-hosting](docs/guide-browser.md#host-the-model-files-yourself).
+are then cached. To serve them yourself, pass `modelHost`: see [self-hosting](docs/guide-browser.md#host-the-model-files-yourself). onnxruntime's wasm comes from jsDelivr unless you pass [`wasmHost`](docs/guide-browser.md#host-the-wasm-yourself).
 
 Before 0.1.1 the default id was `Xenova/bge-small-en-v1.5`. The old ids and heads recording them keep working: they
 resolve to the new ids, so `d.model` returns the new id.

package/dist/index.d.ts

@@ -90,6 +90,12 @@ export interface DeciderOptions {
      * `https://cdn.example.com/models/debuffjp/decide-small/onnx/model_quantized.onnx`. Applies to the whole page or process.
      */
     modelHost?: string;
+    /**
+     * Serve onnxruntime's wasm from your own host instead of jsDelivr, as `<wasmHost>/ort-wasm-simd-threaded.wasm` and
+     * `.mjs` (`ort-wasm-simd-threaded.asyncify.*` with `device: "webgpu"`). Copy them from `onnxruntime-web/dist`.
+     * Applies to the whole page or process.
+     */
+    wasmHost?: string;
     /** Weight precision. Default `q8`. */
     dtype?: "q8" | "fp16" | "fp32";
     /** Called as model files download (first load only; later loads come from cache). */
@@ -109,7 +115,7 @@ export interface DeciderOptions {
  * `train()`. In a browser, prefer `createWorkerDecider()` from `decidekit/browser`: the encoder blocks the thread
  * it runs on.
  */
-export declare function createDecider({ model: requested, device, dtype, temperature, onProgress, modelHost }?: DeciderOptions): Promise<{
+export declare function createDecider({ model: requested, device, dtype, temperature, onProgress, modelHost, wasmHost }?: DeciderOptions): Promise<{
     model: string;
     decide: {
         <const L extends string>(text: string, using: Choices<L>, options?: {

package/dist/index.js

@@ -40,10 +40,20 @@ function toDecision(labels, p) {
     const probabilities = Object.fromEntries(labels.map((l, i) => [l, p[i]]));
     return { choice: labels[best], confidence: p[best], probabilities };
 }
-function usePlainWasm() {
+function chooseWasm(webgpu, wasmHost) {
+    const wasm = env.backends.onnx.wasm;
+    if (!wasm)
+        return;
+    if (wasmHost) {
+        const file = (wasmHost.endsWith("/") ? wasmHost : wasmHost + "/") + (webgpu ? "ort-wasm-simd-threaded.asyncify" : "ort-wasm-simd-threaded");
+        wasm.wasmPaths = { wasm: file + ".wasm", mjs: file + ".mjs" };
+        return;
+    }
+    if (webgpu)
+        return;
     // transformers.js points browsers other than Safari at the 23.6 MB WebGPU-capable build whatever JS is bundled.
     // Leaves paths the app set itself alone.
-    const paths = env.backends.onnx.wasm?.wasmPaths;
+    const paths = wasm.wasmPaths;
     if (!paths || typeof paths === "string" || !paths.wasm?.startsWith("https://cdn.jsdelivr.net/npm/onnxruntime-web@"))
         return;
     paths.wasm = paths.wasm.replace(".asyncify.", ".");
@@ -54,14 +64,13 @@ function usePlainWasm() {
  * `train()`. In a browser, prefer `createWorkerDecider()` from `decidekit/browser`: the encoder blocks the thread
  * it runs on.
  */
-export async function createDecider({ model: requested = SMALL, device, dtype = "q8", temperature, onProgress, modelHost } = {}) {
+export async function createDecider({ model: requested = SMALL, device, dtype = "q8", temperature, onProgress, modelHost, wasmHost } = {}) {
     const model = canonical(requested);
     if (modelHost) {
         env.remoteHost = modelHost.endsWith("/") ? modelHost : modelHost + "/";
         env.remotePathTemplate = "{model}/";
     }
-    if (device !== "webgpu")
-        usePlainWasm();
+    chooseWasm(device === "webgpu", wasmHost);
     const embed = await pipeline("feature-extraction", model, {
         device, dtype, revision: REVISIONS[model],
         progress_callback: onProgress && ((p) => p.status === "progress" && onProgress({ file: p.file, loaded: p.loaded, total: p.total })),

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "decidekit",
-  "version": "0.2.1",
+  "version": "0.2.2",
   "description": "A System One model in the browser: typed, calibrated text decisions on the user's device. Learns from an LLM when it isn't sure.",
   "license": "MIT",
   "repository": "github:debuffjp/decidekit",