Sponsored by AdsCreator.com
Toolify Toolify logo
AI Models
jepa_proof_barlow_twins

ducanhdinh / jepa_proof_barlow_twins

Run on huggingface.co
huggingface.co
Total runs: 0
24-hour runs: 0
7-day runs: 0
30-day runs: 0
Model's Last Updated: June 15 2026

Introduction of jepa_proof_barlow_twins

Model Details of jepa_proof_barlow_twins

ducanhdinh/jepa_proof_barlow_twins

BERT encoder pretrained from scratch với Barlow Twins .

Augmentation strategy

Hai view được tạo bằng span masking độc lập (non-overlapping):

Mô tả
View 1 Câu gốc với các span ngẫu nhiên bị mask
View 2 Câu gốc với các span ngẫu nhiên khác bị mask (không overlap)
Kiến trúc Barlow Twins 
View 1 ──► Encoder (θ) ──► Projector (θ) ──► z1  ──┐
                                                     ├──► Cross-correlation C = Z1ᵀZ2 / N  ──► Loss
View 2 ──► Encoder (θ) ──► Projector (θ) ──► z2  ──┘

Loss = Σ(C_ii - 1)²  +  λ · Σ_{i≠j} C_ij²
Thông số huấn luyện
Tham số Giá trị
Max sequence length 256
Batch size 256
Epochs 10
Learning rate 0.0001
Projector hidden dim 2048
Projector out dim 8192
Off-diagonal coeff (λ) 0.005
Max span length 5
Cách dùng 
from transformers import BertModel, BertTokenizerFast
import torch

tokenizer = BertTokenizerFast.from_pretrained("ducanhdinh/jepa_proof_barlow_twins")
bert      = BertModel.from_pretrained("ducanhdinh/jepa_proof_barlow_twins/encoder")

encoded = tokenizer(
    ["Hello world!", "Barlow Twins pretraining."],
    return_tensors="pt",
    padding=True,
    truncation=True,
)
with torch.no_grad():
    out     = bert(**encoded)
    cls_emb = out.last_hidden_state[:, 0, :]   # [CLS] token → (B, 768)

Runs of ducanhdinh jepa_proof_barlow_twins on huggingface.co

0
Total runs
0
24-hour runs
0
3-day runs
0
7-day runs
0
30-day runs

More Information About jepa_proof_barlow_twins huggingface.co Model

More jepa_proof_barlow_twins license Visit here:

https://choosealicense.com/licenses/apache-2.0

jepa_proof_barlow_twins huggingface.co

jepa_proof_barlow_twins huggingface.co is an AI model on huggingface.co that provides jepa_proof_barlow_twins's model effect (), which can be used instantly with this ducanhdinh jepa_proof_barlow_twins model. huggingface.co supports a free trial of the jepa_proof_barlow_twins model, and also provides paid use of the jepa_proof_barlow_twins. Support call jepa_proof_barlow_twins model through api, including Node.js, Python, http.

jepa_proof_barlow_twins huggingface.co Url

https://huggingface.co/ducanhdinh/jepa_proof_barlow_twins

ducanhdinh jepa_proof_barlow_twins online free

jepa_proof_barlow_twins huggingface.co is an online trial and call api platform, which integrates jepa_proof_barlow_twins's modeling effects, including api services, and provides a free online trial of jepa_proof_barlow_twins, you can try jepa_proof_barlow_twins online for free by clicking the link below.

ducanhdinh jepa_proof_barlow_twins online free url in huggingface.co:

https://huggingface.co/ducanhdinh/jepa_proof_barlow_twins

jepa_proof_barlow_twins install

jepa_proof_barlow_twins is an open source model from GitHub that offers a free installation service, and any user can find jepa_proof_barlow_twins on GitHub to install. At the same time, huggingface.co provides the effect of jepa_proof_barlow_twins install, users can directly use jepa_proof_barlow_twins installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

jepa_proof_barlow_twins install url in huggingface.co:

https://huggingface.co/ducanhdinh/jepa_proof_barlow_twins

Url of jepa_proof_barlow_twins

jepa_proof_barlow_twins huggingface.co Url

Provider of jepa_proof_barlow_twins huggingface.co

ducanhdinh
ORGANIZATIONS
huggingface.co

Other API from ducanhdinh

 huggingface.co
ducanhdinh/HuTieuBert

Total runs: 679
Run Growth: 459
Growth Rate: 67.60%
Updated:November 24 2025