Datasets:
Upload README.md
Browse files
README.md
CHANGED
|
@@ -70,8 +70,8 @@ configs:
|
|
| 70 |
|
| 71 |
| 🌐 Ngôn Ngữ / Language | 📌 Chuyển Hướng / Quick Navigation |
|
| 72 |
| :--- | :--- |
|
| 73 |
-
| **🇻🇳 Bản Tiếng Việt Đầy Đủ** | 👉 [**Xem toàn bộ Báo cáo Kỹ thuật Tiếng Việt**](#vietnamese) |
|
| 74 |
-
| **🇬🇧 Full English Version** | 👉 [**Jump to the complete English Technical Dataset Card**](#english) |
|
| 75 |
|
| 76 |
</div>
|
| 77 |
|
|
@@ -420,7 +420,25 @@ trainer.train()
|
|
| 420 |
|
| 421 |
---
|
| 422 |
|
| 423 |
-
## 8.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 424 |
|
| 425 |
Nếu bạn thấy dự án **Vietverse-SFT** hữu ích và muốn tiếp thêm nguồn lực (chi phí API & Compute) để mở rộng các tập dữ liệu tiếng Việt quy mô lớn hơn trong tương lai, mọi sự ủng hộ đều rất quý báu:
|
| 426 |
|
|
@@ -574,6 +592,7 @@ dataset = load_dataset("TTP01/Vietverse-SFT-Preview", split="train")
|
|
| 574 |
print(f"✅ Loaded {len(dataset)} samples")
|
| 575 |
print("User:", dataset[0]["messages"][0]["content"])
|
| 576 |
print("Assistant:", dataset[0]["messages"][1]["content"])
|
|
|
|
| 577 |
```
|
| 578 |
|
| 579 |
---
|
|
@@ -606,11 +625,11 @@ model = FastLanguageModel.get_peft_model(
|
|
| 606 |
# 3. Format Chat Template
|
| 607 |
dataset = load_dataset("TTP01/Vietverse-SFT-Preview", split="train")
|
| 608 |
|
| 609 |
-
def
|
| 610 |
texts = [tokenizer.apply_chat_template(msgs, tokenize=False, add_generation_prompt=False) for msgs in batch["messages"]]
|
| 611 |
return {"text": texts}
|
| 612 |
|
| 613 |
-
dataset = dataset.map(
|
| 614 |
|
| 615 |
# 4. Train Model
|
| 616 |
trainer = SFTTrainer(
|
|
@@ -766,10 +785,10 @@ trainer.train()
|
|
| 766 |
|
| 767 |
---
|
| 768 |
|
| 769 |
-
## 7.
|
| 770 |
|
| 771 |
-
- **
|
| 772 |
-
- **
|
| 773 |
|
| 774 |
---
|
| 775 |
|
|
|
|
| 70 |
|
| 71 |
| 🌐 Ngôn Ngữ / Language | 📌 Chuyển Hướng / Quick Navigation |
|
| 72 |
| :--- | :--- |
|
| 73 |
+
| **🇻🇳 Bản Tiếng Việt Đầy Đủ (9 Mục)** | 👉 [**Xem toàn bộ Báo cáo Kỹ thuật Tiếng Việt**](#vietnamese) |
|
| 74 |
+
| **🇬🇧 Full English Version (9 Sections)** | 👉 [**Jump to the complete English Technical Dataset Card**](#english) |
|
| 75 |
|
| 76 |
</div>
|
| 77 |
|
|
|
|
| 420 |
|
| 421 |
---
|
| 422 |
|
| 423 |
+
## 8. Giấy Phép & Trích Dẫn (License & Citation)
|
| 424 |
+
|
| 425 |
+
- **Giấy phép**: Phát hành theo giấy phép mở **Apache License 2.0**.
|
| 426 |
+
- **Trích dẫn BibTeX**:
|
| 427 |
+
|
| 428 |
+
```bibtex
|
| 429 |
+
@dataset{vietverse_sft_preview_2026,
|
| 430 |
+
author = {TTP01 and Vietverse Contributors},
|
| 431 |
+
title = {Vietverse-SFT: A High-Density, Native Vietnamese Foundation Dataset for Large Language Model Instruction Tuning},
|
| 432 |
+
year = {2026},
|
| 433 |
+
publisher = {Hugging Face},
|
| 434 |
+
howpublished = {\url{https://huggingface.co/datasets/TTP01/Vietverse-SFT-Preview}},
|
| 435 |
+
version = {1.0-preview}
|
| 436 |
+
}
|
| 437 |
+
```
|
| 438 |
+
|
| 439 |
+
---
|
| 440 |
+
|
| 441 |
+
## 9. Ủng Hộ Dự Án & Liên Hệ (Support & Contact)
|
| 442 |
|
| 443 |
Nếu bạn thấy dự án **Vietverse-SFT** hữu ích và muốn tiếp thêm nguồn lực (chi phí API & Compute) để mở rộng các tập dữ liệu tiếng Việt quy mô lớn hơn trong tương lai, mọi sự ủng hộ đều rất quý báu:
|
| 444 |
|
|
|
|
| 592 |
print(f"✅ Loaded {len(dataset)} samples")
|
| 593 |
print("User:", dataset[0]["messages"][0]["content"])
|
| 594 |
print("Assistant:", dataset[0]["messages"][1]["content"])
|
| 595 |
+
print("Quality Score:", dataset[0]["quality_score"])
|
| 596 |
```
|
| 597 |
|
| 598 |
---
|
|
|
|
| 625 |
# 3. Format Chat Template
|
| 626 |
dataset = load_dataset("TTP01/Vietverse-SFT-Preview", split="train")
|
| 627 |
|
| 628 |
+
def apply_template(batch):
|
| 629 |
texts = [tokenizer.apply_chat_template(msgs, tokenize=False, add_generation_prompt=False) for msgs in batch["messages"]]
|
| 630 |
return {"text": texts}
|
| 631 |
|
| 632 |
+
dataset = dataset.map(apply_template, batched=True)
|
| 633 |
|
| 634 |
# 4. Train Model
|
| 635 |
trainer = SFTTrainer(
|
|
|
|
| 785 |
|
| 786 |
---
|
| 787 |
|
| 788 |
+
## 7. Limitations & Ethical Considerations
|
| 789 |
|
| 790 |
+
- **Synthetic Data Boundary**: While audited through 3-stage multi-agent verification, synthetic generations may occasionally contain minor factual inaccuracies in niche legal citations or dates. Users should cross-reference original primary sources for critical legal or medical decisions.
|
| 791 |
+
- **Safety & PII**: The corpus has been strictly filtered and decontaminated against personally identifiable information (PII), toxic expressions, and sensitive private data.
|
| 792 |
|
| 793 |
---
|
| 794 |
|