Downloads · 30 days
17
3% of all-time downloads
junshengma/Qwythos-9B-WeChat
Qwythos-9B-WeChat is a text generation model from junshengma. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
基于 empero-ai/Qwythos-9B-Claude-Mythos-5-1M(Qwen3.5-9B 架构)LoRA 微调后合并的模型,专用于生成微信公众号科技 / AI / 新闻类文章(Markdown 格式输出)。
Downloads · 30 days
17
3% of all-time downloads
All-time downloads
651
Public
Parameters
9.4B
51.9 GB on disk
Likes
1
Public
Click a slice to open those files.
.safetensors18.9 GB · 100%
From the Hugging Face model README
基于 empero-ai/Qwythos-9B-Claude-Mythos-5-1M(Qwen3.5-9B 架构)LoRA 微调后合并的模型,专用于生成微信公众号科技 / AI / 新闻类文章(Markdown 格式输出)。
## 小标题,训练数据结构化排版占比从 5% → 55%repetition_penalty=1.1,vLLM 加 --reasoning-parser qwen3import torch
from transformers import AutoModelForImageTextToText, AutoTokenizer
model_id = "junshengma/Qwythos-9B-WeChat" # 或本地 D:/models/Qwythos-9B-WeChat-v2
tok = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForImageTextToText.from_pretrained(
model_id, dtype=torch.bfloat16, device_map="auto"
)
msgs = [{"role": "user", "content": (
"你是一位专业的微信公众号文章作者,擅长撰写科技、AI、新闻类文章。"
"请根据给定的主题创作一篇高质量的公众号文章,使用 Markdown 格式输出。\n\n"
"主题:OpenAI 发布最新推理模型,性能大幅提升"
)}]
inputs = tok.apply_chat_template(
msgs, add_generation_prompt=True, tokenize=True,
return_dict=True, return_tensors="pt"
).to(model.device)
out = model.generate(**inputs, max_new_tokens=1500, temperature=0.8, top_p=0.95, repetition_penalty=1.1)
print(tok.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
vllm serve junshengma/Qwythos-9B-WeChat \
--max-model-len 49152 \
--gpu-memory-utilization 0.90 \
--reasoning-parser qwen3
lora-adapter/ 目录内附原始 LoRA adapter(111MB)