Hugging Face's logo

Yougen
/
Qwen3Fangwusha14B

Text Generation
Transformers
Safetensors
Chinese
qwen3
fangwusha
chinese-llm
15b
conversational
text-generation-inference
Model card Files
xet
 Community

Model Card for Yougen/Qwen3Fangwusha14B

Qwen3Fangwusha14Bๆ˜ฏๅŸบไบŽQwen3-14B่ฟ›่กŒๅพฎ่ฐƒ็š„ไธญๆ–‡ๅคง่ฏญ่จ€ๆจกๅž‹๏ผŒไธ“ๆณจไบŽๆๅ‡ไธญๆ–‡ๅฏน่ฏ่ƒฝๅŠ›ใ€ๆŒ‡ไปค้ตๅพชๅ’Œ้€š็”จไปปๅŠก่กจ็Žฐใ€‚่ฏฅๆจกๅž‹ๅฑžไบŽFangwusha็ณปๅˆ—๏ผŒๆ—จๅœจไธบไธญๆ–‡็”จๆˆทๆไพ›้ซ˜่ดจ้‡ใ€ๅฎ‰ๅ…จๅฏ้ ็š„AIๅŠฉๆ‰‹ๆœๅŠกใ€‚

Model Details

Model Description

Qwen3Fangwusha14Bๆ˜ฏไธ€ไธช150ไบฟๅ‚ๆ•ฐ็š„่‡ชๅ›žๅฝ’่ฏญ่จ€ๆจกๅž‹๏ผŒๅœจQwen3-14BๅŸบ็ก€ไธŠ้€š่ฟ‡้ซ˜่ดจ้‡ไธญๆ–‡ๆ•ฐๆฎ้›†่ฟ›่กŒไบ†่ฟ›ไธ€ๆญฅๅพฎ่ฐƒใ€‚ๆจกๅž‹้‡‡็”จBF16็ฒพๅบฆ่ฎญ็ปƒ๏ผŒไผ˜ๅŒ–ไบ†ไธญๆ–‡่ฏญไน‰็†่งฃใ€้€ป่พ‘ๆŽจ็†ๅ’Œๅคš่ฝฎๅฏน่ฏ่ƒฝๅŠ›๏ผŒ้€‚็”จไบŽๅ„็งไธญๆ–‡่‡ช็„ถ่ฏญ่จ€ๅค„็†ไปปๅŠกใ€‚

  • Developed by: Yougen Yuan
  • Funded by [optional]: [More Information Needed]
  • Shared by [optional]: Yougen Yuan
  • Model type: Auto-regressive language model (Decoder-only)
  • Language(s) (NLP): ไธญๆ–‡ (zh), ่‹ฑๆ–‡ (en)
  • License: Apache-2.0
  • Finetuned from model [optional]: Qwen/Qwen3-14B

Model Sources [optional]

Uses

Direct Use

่ฏฅๆจกๅž‹ๅฏ็›ดๆŽฅ็”จไบŽไปฅไธ‹ไปปๅŠก๏ผš

  • ไธญๆ–‡ๅฏน่ฏไธŽ้—ฎ็ญ”
  • ๆ–‡ๆœฌ็”ŸๆˆไธŽ็ปญๅ†™
  • ไฟกๆฏๆๅ–ไธŽๆ€ป็ป“
  • ็ฟป่ฏ‘ไธŽ่ฏญ่จ€่ฝฌๆข
  • ไปฃ็ ่พ…ๅŠฉไธŽ่งฃ้‡Š
  • ๅˆ›ๆ„ๅ†™ไฝœไธŽๅ†…ๅฎนๅˆ›ไฝœ

Downstream Use [optional]

่ฏฅๆจกๅž‹ๅฏ่ฟ›ไธ€ๆญฅๅพฎ่ฐƒ็”จไบŽ๏ผš

  • ็‰นๅฎš้ข†ๅŸŸ็Ÿฅ่ฏ†ๅบ“้—ฎ็ญ”
  • ๅฎขๆˆทๆœๅŠกๆœบๅ™จไบบ
  • ๆ•™่‚ฒ่พ…ๅฏผ็ณป็ปŸ
  • ไผไธšๅ†…้ƒจๆ™บ่ƒฝๅŠฉๆ‰‹
  • ๅ†…ๅฎนๅฎกๆ ธไธŽๅˆ†็ฑป

Out-of-Scope Use

่ฏฅๆจกๅž‹ไธๅบ”็”จไบŽ๏ผš

  • ็”Ÿๆˆ่ฟๆณ•ใ€ๆœ‰ๅฎณใ€ๆšดๅŠ›ๆˆ–ๆญง่ง†ๆ€งๅ†…ๅฎน
  • ๆœช็ปๆŽˆๆƒ็š„ๅŒป็–—่ฏŠๆ–ญใ€ๆณ•ๅพ‹ๅปบ่ฎฎๆˆ–้‡‘่žๆŠ•่ต„ๅปบ่ฎฎ
  • ๅ†’ๅ……ไป–ไบบๆˆ–่ฟ›่กŒๆฌบ่ฏˆๆดปๅŠจ
  • ็”Ÿๆˆๅฏ่ƒฝไพต็Šฏ็Ÿฅ่ฏ†ไบงๆƒ็š„ๅ†…ๅฎน
  • ้ซ˜้ฃŽ้™ฉๅ†ณ็ญ–็ณป็ปŸ๏ผˆๅฆ‚่‡ชๅŠจ้ฉพ้ฉถใ€ๅŒป็–—่ฎพๅค‡ๆŽงๅˆถ็ญ‰๏ผ‰

Bias, Risks, and Limitations

  • ๆจกๅž‹ๅฏ่ƒฝไผš็”Ÿๆˆไธๅ‡†็กฎใ€ไธๅฎŒๆ•ดๆˆ–่ฏฏๅฏผๆ€ง็š„ไฟกๆฏ๏ผŒ็‰นๅˆซๆ˜ฏๅœจๅค„็†ไธ“ไธš้ข†ๅŸŸ็Ÿฅ่ฏ†ๆ—ถ
  • ๆจกๅž‹ๅฏ่ƒฝไผšๅๆ˜ ่ฎญ็ปƒๆ•ฐๆฎไธญๅญ˜ๅœจ็š„ๅ่งๅ’Œๅˆปๆฟๅฐ่ฑก
  • ๆจกๅž‹ๅœจๅค„็†้•ฟๆ–‡ๆœฌๆ—ถๅฏ่ƒฝไผšๅ‡บ็ŽฐไธŠไธ‹ๆ–‡็†่งฃ่ƒฝๅŠ›ไธ‹้™็š„ๆƒ…ๅ†ต
  • ๆจกๅž‹ๅฏ่ƒฝไผšไบง็”Ÿๅนป่ง‰๏ผŒ็ผ–้€ ไธๅญ˜ๅœจ็š„ไบ‹ๅฎžๆˆ–ๅผ•็”จ
  • ๆจกๅž‹็š„่‹ฑๆ–‡่ƒฝๅŠ›็›ธๅฏนไธญๆ–‡่พƒๅผฑ

Recommendations

็”จๆˆทๅœจไฝฟ็”จ่ฏฅๆจกๅž‹ๆ—ถๅบ”๏ผš

  • ๅฏนๆจกๅž‹็”Ÿๆˆ็š„ๅ†…ๅฎน่ฟ›่กŒไบ‹ๅฎžๆ ธๆŸฅๅ’Œ้ชŒ่ฏ
  • ๆ„่ฏ†ๅˆฐๆจกๅž‹ๅฏ่ƒฝๅญ˜ๅœจ็š„ๅ่งๅ’Œๅฑ€้™ๆ€ง
  • ๅœจ้ซ˜้ฃŽ้™ฉๅœบๆ™ฏไธญ่ฐจๆ…Žไฝฟ็”จ๏ผŒๅฟ…่ฆๆ—ถๅ’จ่ฏขไธ“ไธšไบบๅฃซ
  • ้ตๅฎˆ็›ธๅ…ณๆณ•ๅพ‹ๆณ•่ง„ๅ’Œ้“ๅพท่ง„่Œƒ
  • ๆŠฅๅ‘Šไปปไฝ•ๆœ‰ๅฎณๆˆ–ไธๅฝ“็š„ๆจกๅž‹่พ“ๅ‡บ

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_name = "Yougen/Qwen3Fangwusha14B"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

prompt = "ไฝ ๅฅฝ๏ผŒ่ฏทไป‹็ปไธ€ไธ‹ไฝ ่‡ชๅทฑใ€‚"
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

outputs = model.generate(
    **inputs,
    max_new_tokens=512,
    temperature=0.7,
    top_p=0.9,
    do_sample=True
)

response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

Training Details

Training Data

่ฏฅๆจกๅž‹ไฝฟ็”จไบ†ๅคš็ง้ซ˜่ดจ้‡ไธญๆ–‡ๆ•ฐๆฎ้›†่ฟ›่กŒๅพฎ่ฐƒ๏ผŒๅŒ…ๆ‹ฌ๏ผš

  • ้€š็”จๅฏน่ฏๆ•ฐๆฎ้›†
  • ๆŒ‡ไปค้ตๅพชๆ•ฐๆฎ้›†
  • ็Ÿฅ่ฏ†้—ฎ็ญ”ๆ•ฐๆฎ้›†
  • ้€ป่พ‘ๆŽจ็†ๆ•ฐๆฎ้›†

ๆ‰€ๆœ‰ๆ•ฐๆฎ้›†ๅ‡็ป่ฟ‡ไธฅๆ ผ็š„่ดจ้‡่ฟ‡ๆปคๅ’ŒๅŽป้‡ๅค„็†๏ผŒ็กฎไฟ่ฎญ็ปƒๆ•ฐๆฎ็š„่ดจ้‡ๅ’Œๅคšๆ ทๆ€งใ€‚

Training Procedure

Preprocessing [optional]

่ฎญ็ปƒๆ•ฐๆฎ็ป่ฟ‡ไบ†ไปฅไธ‹้ข„ๅค„็†ๆญฅ้ชค๏ผš

  • ๆ–‡ๆœฌๆธ…ๆด—ๅ’Œๆ ‡ๅ‡†ๅŒ–
  • ๆ ผๅผ็ปŸไธ€ๅ’Œ่ง„่ŒƒๅŒ–
  • ่ดจ้‡่ฟ‡ๆปคๅ’ŒๅŽป้‡
  • ๆ•ฐๆฎๅขžๅผบๅ’Œๅคšๆ ทๅŒ–

Training Hyperparameters

  • Training regime: BF16 mixed precision
  • Optimizer: AdamW
  • Learning rate: [More Information Needed]
  • Batch size: [More Information Needed]
  • Epochs: [More Information Needed]
  • Warmup steps: [More Information Needed]
  • Weight decay: [More Information Needed]

Speeds, Sizes, Times [optional]

  • Model size: 15B parameters
  • Checkpoint size: ~30GB (BF16)
  • Training duration: [More Information Needed]
  • Training hardware: [More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

ๆจกๅž‹ๅœจไปฅไธ‹ๅŸบๅ‡†ๆต‹่ฏ•้›†ไธŠ่ฟ›่กŒไบ†่ฏ„ไผฐ๏ผš

  • C-Eval (ไธญๆ–‡้€š็”จ่ƒฝๅŠ›่ฏ„ไผฐ)
  • MMLU (ๅคšไปปๅŠก่ฏญ่จ€็†่งฃ)
  • GSM8K (ๆ•ฐๅญฆๆŽจ็†)
  • HumanEval (ไปฃ็ ็”Ÿๆˆ)

Factors

่ฏ„ไผฐๆถต็›–ไบ†ไปฅไธ‹็ปดๅบฆ๏ผš

  • ็Ÿฅ่ฏ†ๆŽŒๆก็จ‹ๅบฆ
  • ้€ป่พ‘ๆŽจ็†่ƒฝๅŠ›
  • ๆŒ‡ไปค้ตๅพช่ƒฝๅŠ›
  • ไธญๆ–‡็†่งฃไธŽ็”Ÿๆˆ่ƒฝๅŠ›
  • ไปฃ็ ็”Ÿๆˆ่ƒฝๅŠ›

Metrics

  • Accuracy: ็”จไบŽ็Ÿฅ่ฏ†้—ฎ็ญ”ๅ’Œ้€‰ๆ‹ฉ้ข˜ไปปๅŠก
  • Pass@k: ็”จไบŽไปฃ็ ็”ŸๆˆไปปๅŠก
  • BLEU/ROUGE: ็”จไบŽๆ–‡ๆœฌ็”Ÿๆˆๅ’Œ็ฟป่ฏ‘ไปปๅŠก
  • Human evaluation: ็”จไบŽๅฏน่ฏ่ดจ้‡ๅ’Œๆ•ดไฝ“่กจ็Žฐ่ฏ„ไผฐ

Results

[More Information Needed]

Summary

[More Information Needed]

Model Examination [optional]

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

  • Hardware Type: [More Information Needed]
  • Hours used: [More Information Needed]
  • Cloud Provider: [More Information Needed]
  • Compute Region: [More Information Needed]
  • Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

่ฏฅๆจกๅž‹ๅŸบไบŽQwen3ๆžถๆž„๏ผŒ้‡‡็”จ่งฃ็ ๅ™จ-only็š„Transformer็ป“ๆž„๏ผš

  • ไธŠไธ‹ๆ–‡็ช—ๅฃๅคงๅฐ๏ผš[More Information Needed]
  • ๆณจๆ„ๅŠ›ๆœบๅˆถ๏ผšGrouped-Query Attention (GQA)
  • ๆฟ€ๆดปๅ‡ฝๆ•ฐ๏ผšSwiGLU
  • ่ฏ่กจๅคงๅฐ๏ผš[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

  • Framework: PyTorch 2.x
  • Training library: LLaMA-Factory
  • Inference library: Transformers 4.x
  • Acceleration: FlashAttention-2

Citation [optional]

BibTeX:

@misc{qwen3fangwusha14b,
  author = {Yuan, Yougen},
  title = {Qwen3Fangwusha14B: A Fine-tuned Chinese Large Language Model},
  year = {2026},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/Yougen/Qwen3Fangwusha14B}}
}

APA:

Yuan, Y. (2026). Qwen3Fangwusha14B: A Fine-tuned Chinese Large Language Model. Hugging Face. https://huggingface.co/Yougen/Qwen3Fangwusha14B

Glossary [optional]

[More Information Needed]

More Information [optional]

่ฏฅๆจกๅž‹ๆ˜ฏFangwusha็ณปๅˆ—็š„ไธ€้ƒจๅˆ†๏ผŒๆ›ดๅคš็›ธๅ…ณๆจกๅž‹ๅฏๅœจไปฅไธ‹้›†ๅˆไธญๆ‰พๅˆฐ๏ผš

Model Card Authors [optional]

Yougen Yuan

Model Card Contact

[More Information Needed]

Downloads last month
2
Safetensors
Model size
15B params
Tensor type
BF16
ยท
Inference Providers NEW
Text Generation
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for Yougen/Qwen3Fangwusha14B

Base model

Qwen/Qwen3-14B-Base
Finetuned
Qwen/Qwen3-14B
Finetuned
(387)
this model

Collection including Yougen/Qwen3Fangwusha14B