GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
rubenich
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
rubenich
Model Tree
Input Modalities
Output Modalities
Supported Functionality
LlamaForCausalLM): hidden 256, 6 layers,
8 heads, intermediate 1024[BOS]/[EOS])[PAD] [UNK] [BOS] [EOS] A C G T N. One base = one token,
so a single-base change alters exactly one token. Lowercase (soft-masked) input is
uppercased and any other letter becomes N.hgdownload.soe.ucsc.eduN)Bits per base on chr21 (35,087,360 bases in windows without N),
each base predicted from the bases before it in its window. Uniform guessing = 2.0;
lower is better.
| Method | Bits per base |
|---|---|
| This model | 1.7439 |
| Markov order 0 | 1.9905 |
| Markov order 1 | 1.9522 |
| Markov order 2 | 1.9383 |
| Markov order 3 | 1.9326 |
| Markov order 4 | 1.9258 |
| Markov order 5 | 1.9178 |
| Markov order 6 | 1.9014 |
| Markov order 8 | 1.8574 |
| Markov order 10 | 1.8308 |
| Uniform | 2.0000 |
python
import torchfrom transformers import AutoModelForCausalLM, AutoTokenizer tok = AutoTokenizer.from_pretrained("rubenich/genetic-dna-tiny")model = AutoModelForCausalLM.from_pretrained("rubenich/genetic-dna-tiny") ids = tok("ACGTTGCA", return_tensors="pt")["input_ids"]with torch.no_grad(): logits = model(input_ids=ids).logitsZero-shot variant scoring compares log P(alt window) - log P(ref window) on an hg19
window around a single-base change; see genetic_model.score_variant in the source repo.
import torchfrom transformers import AutoModelForCausalLM, AutoTokenizer tok = AutoTokenizer.from_pretrained("rubenich/genetic-dna-tiny")model = AutoModelForCausalLM.from_pretrained("rubenich/genetic-dna-tiny") ids = tok("ACGTTGCA", return_tensors="pt")["input_ids"]with torch.no_grad(): logits = model(input_ids=ids).logits