Downloads · 30 days
0
BuerM/aqr
aqr is a machine learning model from BuerM. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Selected checkpoints for the main benchmark table and key controls. Code lives in the git repo; this Hub repo stores inference weights only.
Downloads · 30 days
0
Access
Public
Updated Sep 23, 2026
Repo size
9.6 GB
Likes
1
Public
Click a slice to open those files.
.pt9.2 GB · 96%
From the Hugging Face model README
Selected checkpoints for the main benchmark table and key controls. Code lives in the git repo; this Hub repo stores inference weights only.
| Setting | Step | Path |
|---|---|---|
| Qwen3-0.6B ← Qwen2.5-0.5B-Instruct | 6500 | qcomm/controlled/checkpoint-6500 |
| Qwen3-0.6B ← Qwen3-4B-Base | 6000 | qcomm/qwen3-4b-base/checkpoint-6000 |
| Qwen3-0.6B ← Llama-3.2-1B-Instruct | 3000 | qcomm/llama3.2-1b/checkpoint-3000 |
Selection uses held-out HellaSwag (checkpoint_selection.json). SHA-256 manifests are under checkpoint_manifests/.
baselines/published_c2c/ — official C2C fusers used in the main table (also available at nics-efc/C2C_Fuser)baselines/c2c_matched/checkpoint-500 — protocol-matched C2C retrainbaselines/cross_attn_adapter/checkpoint-500 — parameter-matched cross-attention adapterbaselines/receiver_only_adapter/checkpoint-500 — receiver-only adapterhf download BuerM/aqr --local-dir ./aqr_ckpts