From 2a7e8e776a6004fd1f144766b3ca7d00be0a9e71 Mon Sep 17 00:00:00 2001 From: IC Date: Wed, 19 Aug 2026 15:03:39 +0000 Subject: [PATCH 1/2] chore: align @qvac/sdk addon pins on the fabric-10069.1.1 releases Raise the SDK's native addon floors to the releases that carry qvac-fabric 10069.1.1 (Adreno 830 OpenCL MoE repack fix, GPU MoE kernels re-enabled), so the flattened mobile lib dir ships one consistent GGML backend set again: - @qvac/classification-ggml ^0.18.0 -> ^0.20.0 (brings @qvac/fabric 0.6.0) - @qvac/embed-llamacpp ^0.32.0 -> ^0.34.0 - @qvac/llm-llamacpp ^0.43.0 -> ^0.45.0 - @qvac/ocr-ggml ^0.16.0 -> ^0.18.0 - @qvac/translation-nmtcpp ^0.8.0 -> ^0.10.0 - @qvac/vla-ggml ^0.19.0 -> ^0.21.0 All target versions are published and anonymously reachable on the public npm registry. Every changelog entry in these ranges is a fabric bump with no API change, apart from llm-llamacpp 0.44.0's additive image_no_upscale load option, so no SDK code changes are required. --- packages/sdk/package.json | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/packages/sdk/package.json b/packages/sdk/package.json index 6dc0f333af..42ef95f9d9 100644 --- a/packages/sdk/package.json +++ b/packages/sdk/package.json @@ -209,20 +209,20 @@ "@qvac/audiogen-ggml": "^0.2.1", "@qvac/asr-ggml": "^0.3.0", "@qvac/bci-whispercpp": "^0.7.1", - "@qvac/classification-ggml": "^0.18.0", + "@qvac/classification-ggml": "^0.20.0", "@qvac/decoder-audio": "^0.5.0", "@qvac/diffusion-cpp": "^0.17.0", - "@qvac/embed-llamacpp": "^0.32.0", + "@qvac/embed-llamacpp": "^0.34.0", "@qvac/error": "^0.1.1", "@qvac/langdetect-text": "^0.1.2", - "@qvac/llm-llamacpp": "^0.43.0", + "@qvac/llm-llamacpp": "^0.45.0", "@qvac/logging": "^0.1.0", - "@qvac/ocr-ggml": "^0.16.0", + "@qvac/ocr-ggml": "^0.18.0", "@qvac/rag": "^0.6.4", "@qvac/registry-client": "^0.6.1", - "@qvac/translation-nmtcpp": "^0.8.0", + "@qvac/translation-nmtcpp": "^0.10.0", "@qvac/tts-ggml": "^0.7.0", - "@qvac/vla-ggml": "^0.19.0", + "@qvac/vla-ggml": "^0.21.0", "bare-abort-controller": "^1.0.0", "bare-cpu-info": "0.1.1", "bare-crypto": "^1.15.0", From 9eff069a9177c88903776243423c7d035d5b771a Mon Sep 17 00:00:00 2001 From: IC Date: Wed, 19 Aug 2026 17:21:11 +0000 Subject: [PATCH 2/2] chore: raise the @qvac/vla-ggml floor to ^0.21.1 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ^0.21.0 already resolved 0.21.1, but only for a fresh install — a consumer with a lockfile pinning 0.21.0 stays there. 0.21.1 is the CPU SmolVLA mmap fix: the weights path read the device off the buffer type, and ggml declares the CPU buffer type with no device attached, so every CPU load allocated a private copy of the model instead of mapping it and charged the whole file (~1.9 GB for the q8 LIBERO checkpoint) to the process. On iOS that pushed the process against its memory limit and failed the load with a bare "Failed to load SmolVLA model", intermittently and more often later in a long run. Making it the floor means no consumer of this SDK can resolve a vla-ggml without the fix. Verified 0.21.1 is published and anonymously reachable on registry.npmjs.org. --- packages/sdk/package.json | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/packages/sdk/package.json b/packages/sdk/package.json index 615e7a7cf6..d2cb8807e8 100644 --- a/packages/sdk/package.json +++ b/packages/sdk/package.json @@ -222,7 +222,7 @@ "@qvac/registry-client": "^0.6.1", "@qvac/translation-nmtcpp": "^0.10.0", "@qvac/tts-ggml": "^0.7.4", - "@qvac/vla-ggml": "^0.21.0", + "@qvac/vla-ggml": "^0.21.1", "bare-abort-controller": "^1.0.0", "bare-cpu-info": "0.1.1", "bare-crypto": "^1.15.0",