Downloads · 30 days
0
Cortex_A_0.5
PRATYUSH-BHARDWAJ/Cortex_A_0.5
Cortex_A_0.5 is a machine learning model from PRATYUSH-BHARDWAJ. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.
General-purpose edge checkpoint: Qwen3.5-0.8B full SFT with Unsloth int8-int4 QAT (4-bit weights + 8-bit dynamic activations). Target inference footprint ≈ 450MB including the vision tower.
Access
Public
Updated Aug 26, 2026
Repo size
31.4 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt5 GB · 74%
At a glance
- Library
- transformers
- License
- apache-2.0
- Access
- Public
- Created
- Aug 25, 2026
- Updated
- Aug 26, 2026
- SHA
- f6c24e02
Base models
Model card
From the Hugging Face model README
Cortex A 0.5
General-purpose edge checkpoint: Qwen3.5-0.8B full SFT with Unsloth int8-int4 QAT (4-bit weights + 8-bit dynamic activations). Target inference footprint ≈ 450MB including the vision tower.
This repo stores:
checkpoint-*— resumable Trainer states (optimizer + fake-quant QAT model)training/live_metrics.json— loss, MTP loss, ppl, val loss/ppl, tok/s, grad norm, lrtraining/RESUME_POINTER.json— last step for the next 12h Kaggle sessionqat_converted/— real 4-bit TorchAO export (only after a completed epoch run)
Training hardware: Kaggle 2× Tesla T4, hard stop 11.5h, DDP via torchrun.
QAT scheme: int8-int4.
- Library
- transformers
- License
- apache-2.0
- Created
- Aug 25, 2026
- Updated
- Aug 26, 2026