leyoai-video-mini

LeyoAI Video Safety Assistant (Mini)

English | 中文

English

Model Description

Lightweight video content safety classification assistant

Installation

pip install transformers peft torch

Usage Scenarios

• Video content classification
• Safety detection
• Content moderation

How to Use

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

# Load base model
base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-0.5B-Instruct")
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-0.5B-Instruct")

# Load LoRA adapter
model = PeftModel.from_pretrained(base_model, "FFZwai/leyoai-video-mini")

# Generate response
input_text = "Your question here"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs, max_length=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training Details

• Base Model: Qwen/Qwen2.5-0.5B-Instruct
• LoRA Rank: 8
• Training Data: 8,049 samples
• Validation Data: 1,726 samples
• Framework: QLoRA + SFTTrainer

---

中文

模型描述

轻量级视频内容安全分类助手

安装方法

pip install transformers peft torch

使用场景

• 视频内容分类
• 安全检测
• 内容审核

使用方法

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

# 加载基座模型
base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-0.5B-Instruct")
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-0.5B-Instruct")

# 加载 LoRA 适配器
model = PeftModel.from_pretrained(base_model, "FFZwai/leyoai-video-mini")

# 生成回复
input_text = "你的问题"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs, max_length=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

训练细节

• 基座模型: Qwen/Qwen2.5-0.5B-Instruct
• LoRA Rank: 8
• 训练数据: 8,049 samples
• 验证数据: 1,726 samples
• 训练框架: QLoRA + SFTTrainer

---

About LeyoAI

LeyoAI is an AI MaaS platform by 杭州市上城区乐友信息服务工作室, providing specialized AI assistants for cybersecurity, video safety, workflow automation, and data analytics.

Website: https://leyoai.vercel.app