---
license: mit
tags:
- embedding
- text-embedding
- crypto
- nlp
library_name: transformers
---

# crypto-mini-embed

**crypto-mini-embed** adalah contoh model mini embedding berbasis arsitektur sederhana untuk eksperimen NLP seperti:

- text similarity  
- vector search  
- clustering  
- semantic tagging  
- crypto-topic classification  

Model ini merupakan **dummy model** untuk membantu pengguna memahami struktur repository model di HuggingFace.

---

## ⚙️ Arsitektur Model

- Tipe model: `MiniEmbeddingModel`
- Hidden size: 64
- Max length: 128 tokens
- Framework: PyTorch
- Format: Safetensors
- Tokenizer: Basic CharTokenizer (dummy)

---

## 📦 File dalam Model

| File | Fungsi |
|------|--------|
| `config.json` | Konfigurasi model |
| `tokenizer.json` | Tokenizer sederhana |
| `model.safetensors` | Parameter model |
| `README.md` | Dokumentasi model |

---

## 🧪 Contoh Penggunaan

```python
from transformers import AutoTokenizer, AutoModel
import torch

tok = AutoTokenizer.from_pretrained("0xcubin/crypto-mini-embed")
model = AutoModel.from_pretrained("0xcubin/crypto-mini-embed")

text = "Bitcoin is digital money"
inputs = tok(text, return_tensors="pt")

with torch.no_grad():
    emb = model(**inputs).last_hidden_state.mean(dim=1)

print(emb.shape)  # contoh: (1, 64)