anyformer 0.4.3 → 0.4.4

package/CHANGELOG.md

@@ -1,5 +1,17 @@
 # Changelog
 
+## 0.4.4 — 2026-10-24
+
+### Changed
+- diskformer pinned to v0.6.7 (663b2b0). Its hand engines' ingest (src/ingest.js) now streams the GGUF header with
+  progress, so a slow link no longer goes silent for minutes (Flint at 0.1 MB/s stalled LocalMind's watchdog for 3 min).
+  A range that drops mid-stream, or returns 5xx or 429, resumes from the last byte, up to 6 times. The new events reach
+  `onProgress` as status `weights` with phase `header` or `retry`. diskformer changed only src/ingest.js and its test
+  between v0.6.6 and v0.6.7.
+
+### Verified
+- npm test 87/87. Not run here in a browser: LocalMind's throttled production load of Flint is that check.
+
 ## 0.4.3 — 2026-10-24
 
 npm skips 0.4.2: this release carries it (Flint 40B and LFM2.5 24B on the hand engines).

package/README.md

@@ -14,7 +14,7 @@ browser storage; no account, no server, no telemetry.
 | How | Command |
 |---|---|
 | npm | `npm install anyformer` |
-| CDN | `import { load } from 'https://cdn.jsdelivr.net/npm/anyformer@0.4.3/+esm';` |
+| CDN | `import { load } from 'https://cdn.jsdelivr.net/npm/anyformer@0.4.4/+esm';` |
 | Copy | `index.js` and `src/`; nothing else is needed |
 
 ```js

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "anyformer",
-  "version": "0.4.3",
+  "version": "0.4.4",
   "description": "One call to run a model in a browser tab: anyformer reads the Hub repo, probes the browser, and picks the back end (transformers.js, LocalMind's WebGPU kernels, diskformer, kohra) that fits the model's format, architecture and your GPU budget.",
   "type": "module",
   "main": "./index.js",

package/src/sources.js

@@ -7,9 +7,10 @@ export const PINS = {
   transformers: '4.3.1',
   wllama: '3.4.1',      // LocalMind's pin; its stream-drain workaround (src/backends/wllama.js) is for this version
   kohra: 'f678fac7342c79b58b640540b8b3a7d078b7ab1a',
-  // v0.6.6: every engine (the disk-tier MoE engines, Ling 3.0 Flash, and LocalMind's whole-model engines LFM2.5,
+  // v0.6.7: every engine (the disk-tier MoE engines, Ling 3.0 Flash, and LocalMind's whole-model engines LFM2.5,
   // Gemma 4 E2B/E4B and Ternary Bonsai 2) under engines/, and the llama.cpp path under ggml/, which loads split GGUFs.
-  diskformer: 'e07e12975df054b073672ba2060d0d2c5571dbcd',
+  // The hand engines' ingest streams the GGUF header with progress and resumes a dropped range.
+  diskformer: '663b2b01cfbed95ae8d319ce845d8aea759d60cf',
 };
 
 const gh = (repo, sha, path) => `https://cdn.jsdelivr.net/gh/NakliTechie/${repo}@${sha}/${path}`;