Instructions to use mikuhhn1239/qwen3-8b-novel-base-sft with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mikuhhn1239/qwen3-8b-novel-base-sft with Transformers:

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

pipe = pipeline("text-generation", model="mikuhhn1239/qwen3-8b-novel-base-sft")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mikuhhn1239/qwen3-8b-novel-base-sft")
model = AutoModelForCausalLM.from_pretrained("mikuhhn1239/qwen3-8b-novel-base-sft", device_map="auto")
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 Settings

vLLM

How to use mikuhhn1239/qwen3-8b-novel-base-sft with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mikuhhn1239/qwen3-8b-novel-base-sft"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mikuhhn1239/qwen3-8b-novel-base-sft",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mikuhhn1239/qwen3-8b-novel-base-sft

SGLang

How to use mikuhhn1239/qwen3-8b-novel-base-sft 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 "mikuhhn1239/qwen3-8b-novel-base-sft" \
    --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": "mikuhhn1239/qwen3-8b-novel-base-sft",
		"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 "mikuhhn1239/qwen3-8b-novel-base-sft" \
        --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": "mikuhhn1239/qwen3-8b-novel-base-sft",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mikuhhn1239/qwen3-8b-novel-base-sft with Docker Model Runner:
```
docker model run hf.co/mikuhhn1239/qwen3-8b-novel-base-sft
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Qwen3-8B Novel Base SFT

基于 Qwen3-8B 全参微调，用 669 本中文网络小说训练（约 7200 万字符）。

覆盖类型：言情、耽美、都市言情、玄幻、网游。

项目地址：https://github.com/lin1753/novel2galgame

训练代码仓库：https://github.com/lin1753/novel-agent

用途

All Novel Can Be Galgame 工作台的 Stage1 基座模型。

学习中文小说的叙事风格和角色对话模式，作为下游 LoRA adapter 的基座，执行三类 Agent 任务：

Agent 1: mikuhhn1239/qwen3-8b-narrative-parsing-lora — 叙事单元分类
Agent 2: mikuhhn1239/qwen3-8b-scene-segmentation-lora — 场景边界检测
Agent 3: mikuhhn1239/qwen3-8b-attribution-assist-lora — 角色归因

加载

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
    "mikuhhn1239/qwen3-8b-novel-base-sft",
    torch_dtype="auto",
    device_map="auto",
)
tokenizer = AutoTokenizer.from_pretrained(
    "mikuhhn1239/qwen3-8b-novel-base-sft"
)

4-bit 量化（适用于 8GB 显存）：

from transformers import BitsAndBytesConfig

bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_compute_dtype="float16",
    bnb_4bit_quant_type="nf4",
)
model = AutoModelForCausalLM.from_pretrained(
    "mikuhhn1239/qwen3-8b-novel-base-sft",
    quantization_config=bnb_config,
    device_map="auto",
)

训练数据

文件	条数	说明
`continuation.jsonl`	36,092	续写：给前半段→续后半段
`instruction.jsonl`	36,481	指令式续写
合计	72,573

格式: ChatML [system, user, assistant]

训练详情

调试历程

全参微调 8B 模型内存压力大，经历多轮调试：

#	问题	原因	解决
1	单卡 OOM (78G/80G)	optimizer+grads+model≈116G/卡	上 4 卡 DDP
2	4 卡 DDP 仍 OOM (79G/80G)	DDP 每卡存完整 AdamW 状态(66G)	加 DeepSpeed ZeRO-2
3	ZeRO-2 backward OOM	内存碎片	`PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True`
4	batch=2 预估 26 小时	序列太长	seq_len 4096→2048, epochs 3→2, batch 2→4

最终超参

参数	值
方法	全参数 SFT
有效 batch size	64 (4 GPU × 4 batch × 4 accumulation)
学习率	2e-5
优化器	AdamW (adamw_torch_fused)
warmup	3%
scheduler	linear decay
精度	bf16
最大序列长度	2048
epochs	2
gradient checkpointing	True
分布式	DeepSpeed ZeRO-2 (4×A800 80GB)