Models build with 0.3.4, improved quality and bug fixes
ANEMLL: Open Source project (pronounced as animal) PRO
anemll
AI & ML interests
Apple Neural Engine, on-device-compute
Recent Activity
liked
a model
1 day ago
PKU-DS-LAB/Fairy-plus-minus-i-1.3B
liked
a Space
about 1 month ago
muellerzr/gpu-tflop-finder
published
an
article
2 months ago
"Anemll-style" Root-Mean-Square (RMS) Normalization on the Apple Neural Engine: A Simple Hack
Organizations
Just an EZ way to collect papers on HF
-
Reducing Transformer Key-Value Cache Size with Cross-Layer Attention
Paper • 2405.12981 • Published • 33 -
TinyR1-32B-Preview: Boosting Accuracy with Branch-Merge Distillation
Paper • 2503.04872 • Published • 15 -
FFN Fusion: Rethinking Sequential Computation in Large Language Models
Paper • 2503.18908 • Published • 19
conversions done with 0.1.2
Models build with 0.3.4, improved quality and bug fixes
Initial Support for QWEN3
Just an EZ way to collect papers on HF
-
Reducing Transformer Key-Value Cache Size with Cross-Layer Attention
Paper • 2405.12981 • Published • 33 -
TinyR1-32B-Preview: Boosting Accuracy with Branch-Merge Distillation
Paper • 2503.04872 • Published • 15 -
FFN Fusion: Rethinking Sequential Computation in Large Language Models
Paper • 2503.18908 • Published • 19
Unzipped models for iOS downloads, test only
conversions done with 0.1.2
Models converted with ANEMLL - 0.1.11 and one shot convert script https://github.com/Anemll/Anemll/blob/main/docs/convert_model.md