Downloads · 30 days
12
15% of all-time downloads
stillmeta/qwen25-7b-police-lora
qwen25-7b-police-lora is a machine learning model from stillmeta. 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.
Хяналтын камерын бичлэгнээс зодооны зөрчил илрүүлж, Монгол хэлээр тайлбар гаргах зориулалттай fine-tuned загвар.
Downloads · 30 days
12
15% of all-time downloads
All-time downloads
79
Public
Repo size
2.9 GB
Likes
0
Public
Click a slice to open those files.
.pt1.9 GB · 62%
From the Hugging Face model README
Хяналтын камерын бичлэгнээс зодооны зөрчил илрүүлж, Монгол хэлээр тайлбар гаргах зориулалттай fine-tuned загвар.
Qwen3-VL-30B (Teacher)
↓
2,260 хяналтын камерын тайлан
↓
~9,040 Q&A датасет
↓
Qwen2.5-7B LoRA Fine-tuning (Student)
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
from peft import PeftModel
import torch
base_model = "Qwen/Qwen2.5-7B-Instruct"
lora_model = "stillmeta/qwen25-7b-police-lora"
tokenizer = AutoTokenizer.from_pretrained(base_model)
bnb = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_compute_dtype=torch.float16)
model = AutoModelForCausalLM.from_pretrained(base_model, quantization_config=bnb, device_map="auto")
model = PeftModel.from_pretrained(model, lora_model)
model.eval()
| Параметр | Утга |
|---|---|
| LoRA rank | 16 |
| LoRA alpha | 32 |
| Dropout | 0.05 |
| Learning rate | 2e-4 |
| Batch size | 16 (4x4) |
| Epochs | 10 |
| Quantization | 4-bit NF4 |
Батсүхын Бат-Эрдэнэ — ШУТИС, МХТС, Хиймэл оюун ухааны тэнхим, 2026