Instructions to use god520/ZFT-Community-8B-BS4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use god520/ZFT-Community-8B-BS4 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="god520/ZFT-Community-8B-BS4")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("god520/ZFT-Community-8B-BS4")
model = AutoModelForCausalLM.from_pretrained("god520/ZFT-Community-8B-BS4")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use god520/ZFT-Community-8B-BS4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "god520/ZFT-Community-8B-BS4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "god520/ZFT-Community-8B-BS4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/god520/ZFT-Community-8B-BS4

SGLang

How to use god520/ZFT-Community-8B-BS4 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "god520/ZFT-Community-8B-BS4" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "god520/ZFT-Community-8B-BS4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "god520/ZFT-Community-8B-BS4" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "god520/ZFT-Community-8B-BS4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use god520/ZFT-Community-8B-BS4 with Docker Model Runner:
```
docker model run hf.co/god520/ZFT-Community-8B-BS4
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

ZFT-Community-8B-BS4

模型简介 (Model Overview) ZFT-Community-8B-BS4 是一个开源的大型语言模型，基于 Transformer 架构，拥有 80 亿参数。该模型在中文和英文的混合语料上进行了预训练，旨在支持通用自然语言处理任务，如文本生成、对话系统、问答系统、情感分析等。

作者: god520

模型大小: 8B 参数

许可证: [MIT License]

框架: PyTorch + Hugging Face Transformers

权重格式: safetensors

模型用途 (Intended Use)

✅ 推荐用途自然语言生成（NLG）

智能对话系统

文本摘要、扩写

问答系统

情感分类

❌ 非推荐用途生成虚假新闻或误导性信息

自动化决策的法律、医疗场景

任何违反法律法规的用途

模型架构 (Architecture) 类型: 基于 Transformer 的自回归语言模型

参数量: 8B

Tokenizer: Qwen tokenizer

训练框架: PyTorch + Hugging Face Accelerate

文件格式: Safetensors

训练数据 (Training Data) 使用多源公开数据集（包括通用网页文本、百科内容、新闻数据）

总规模：约 200GB 干净文本

数据清理策略：去重、过滤低质量文本、字符规范化

不包含用户隐私数据

性能评估 (Evaluation) 该模型尚未经过 Open-LLM Leaderboard 官方评测，以下为开发者自测结果（如有）：

Perplexity (PPL): 10.5

文本生成: 高可读性，对中文和英文均有较好表现

已测试任务: 问答、摘要、翻译、话题推荐（仅小规模实验）

👉 未来计划: 提交至 Open-LLM Leaderboard，支持标准基准（MMLU、ARC、TruthfulQA、HellaSwag 等）。

局限性 (Limitations) 对专业领域（医学、法律等）表现较弱

在长文本推理和逻辑一致性方面有改进空间

多语言支持有限，中文/英文效果最佳

偏差与伦理 (Bias and Ethical Considerations) 可能存在训练语料带来的隐性偏差

不应将模型用于可能带来伤害的场景

需结合人工审核，避免模型输出敏感或有害内容

许可证 (License) MIT License

用户可自由使用、修改、分发模型，但需遵守相关法律法规，不得用于恶意用途

使用方法 (How to Use) python 复制编辑 from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "god520/ZFT-Community-8B-BS4"

加载模型

tokenizer = AutoTokenizer.from_pretrained(model_name) model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto")

推理示例

input_text = "介绍一下大语言模型的原理。" inputs = tokenizer(input_text, return_tensors="pt").to(model.device) outputs = model.generate(**inputs, max_length=200)

print(tokenizer.decode(outputs[0], skip_special_tokens=True)) 引用 (Citation) 如果你使用本模型，请引用：

复制编辑 @misc{zftcommunity8b, title = {ZFT-Community-8B-BS4: An Open-Source 8B Parameter Language Model}, author = {huazhixin}, year = {2025}, url = {https://huggingface.co/god520/ZFT-Community-8B-BS4} }

✅ 下一步：

将此模型卡复制到 Hugging Face 项目的 README.md 或 Model Card 部分。

必须补充 Open-LLM Leaderboard 要求的字段：Model Name, Organization, License, Parameter Count, Evaluation Results，并附带 Hugging Face 链接。

是否需要我帮你直接生成符合 Open-LLM Leaderboard 官方模板的 YAML/Markdown 提交文件？（可以直接上传到 submission 格式）

Downloads last month: 3

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for god520/ZFT-Community-8B-BS4

Base model

Qwen/Qwen3-8B-Base

Finetuned

(418)

this model

Quantizations

2 models

god520
/

ZFT-Community-8B-BS4

ZFT-Community-8B-BS4

加载模型

推理示例

Model tree for god520/ZFT-Community-8B-BS4

Dataset used to train god520/ZFT-Community-8B-BS4