Downloads · 30 days
0
explcre/carbon-3b-lora-teachers-nt18
carbon-3b-lora-teachers-nt18 is a machine learning model from explcre. 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 peft. The card lists the license as apache-2.0.
Per-task LoRA adapters (PEFT, r=16, alpha=32) fine-tuning HuggingFaceBio/Carbon-3B on the NT-revised 18-task genomics benchmark. Stage-1 teachers for BPNet distillation. Mean test MCC ~ 0.664.
Downloads · 30 days
0
Access
Public
Updated Oct 2, 2026
Repo size
1.9 GB
Likes
0
Public
Click a slice to open those files.
.safetensors1.9 GB · 87%
From the Hugging Face model README
Per-task LoRA adapters (PEFT, r=16, alpha=32) fine-tuning HuggingFaceBio/Carbon-3B on the
NT-revised 18-task genomics benchmark. Stage-1 teachers for BPNet distillation. Mean test MCC ~ 0.664.
Load: base HuggingFaceBio/Carbon-3B + PeftModel.from_pretrained(base, '<task>_finetuned').
| task | test MCC |
|---|---|
| H2AFZ | 0.5138934004275322 |
| H3K27ac | 0.4086581555471983 |
| H3K27me3 | 0.6026541511968219 |
| H3K36me3 | 0.608295752304769 |
| H3K4me1 | 0.47577999488021877 |
| H3K4me2 | 0.5856004186748309 |
| H3K4me3 | 0.6532783808219426 |
| H3K9ac | 0.5358576368989549 |
| H3K9me3 | 0.3502282641336323 |
| H4K20me1 | 0.6711769444959546 |
| enhancers | 0.5369466495428454 |
| enhancers_types | 0.5492346819627704 |
| promoter_all | 0.7745296713025749 |
| promoter_no_tata | 0.7818805257369318 |
| promoter_tata | 0.9717413564871221 |
| splice_sites_acceptors | 0.9733446058823734 |
| splice_sites_all | 0.9720178517017777 |
| splice_sites_donors | 0.9780025010870148 |
These adapters are derived from Carbon-3B, which is released under Apache-2.0.