Note b2 (beta3-loadable) bundles under gpu-pipelined-b2/
Browse files
README.md
CHANGED
|
@@ -14,6 +14,14 @@ tags:
|
|
| 14 |
|
| 15 |
# Youtu-LLM-2B — Apple Core AI (`.aimodel`)
|
| 16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
[Youtu-LLM-2B](https://huggingface.co/tencent/Youtu-LLM-2B) (Tencent) converted to Apple
|
| 18 |
**Core AI** for iOS 27 / macOS 27 (beta) — the **[zoo](https://github.com/john-rocky/coreai-model-zoo)'s
|
| 19 |
first Multi-head Latent Attention (MLA) model that runs on iPhone**, and its first **dense**
|
|
|
|
| 14 |
|
| 15 |
# Youtu-LLM-2B — Apple Core AI (`.aimodel`)
|
| 16 |
|
| 17 |
+
> [!NOTE]
|
| 18 |
+
> **Update 2026-07-15:** `gpu-pipelined-b2/` adds `youtu_llm_2b_decode_absorbed_int8_msdpa_g32` re-exported with
|
| 19 |
+
> `coreai-core 1.0.0b2`, loadable on the OS 27 **beta 3** toolchain (June-era b1
|
| 20 |
+
> bundles fail to load there with a versioned-IR error). The original b1 tree is
|
| 21 |
+
> retained unchanged so existing apps and pinned catalogs keep working. The b2
|
| 22 |
+
> decode bundle is the exact artifact measured on [DeviceMark](https://devicemark.github.io/).
|
| 23 |
+
|
| 24 |
+
|
| 25 |
[Youtu-LLM-2B](https://huggingface.co/tencent/Youtu-LLM-2B) (Tencent) converted to Apple
|
| 26 |
**Core AI** for iOS 27 / macOS 27 (beta) — the **[zoo](https://github.com/john-rocky/coreai-model-zoo)'s
|
| 27 |
first Multi-head Latent Attention (MLA) model that runs on iPhone**, and its first **dense**
|