Downloads · 30 days
6
0% of all-time downloads
CPRetriever-qwen-0.6B-distilled
abcba909/CPRetriever-qwen-0.6B-distilled
CPRetriever-qwen-0.6B-distilled is a machine learning model from abcba909. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as cc-by-nc-4.0.
This is a distilled version of the CPRetriever-Prob-Qwen3-4B-2510 model, obtained through representation alignment distillation on the Qwen3-embedding-0.6B backbone.
All-time downloads
2.4K
Public
Parameters
598M
2.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors2.4 GB · 99%
At a glance
- License
- cc-by-nc-4.0
- Model type
- cp_embedding
- Access
- Public
- Created
- Feb 7, 2026
- Updated
- Mar 26, 2026
- SHA
- 7f9c4ebe
Base models
Model card
From the Hugging Face model README
CPRetriever-qwen-0.6B-distilled
This is a distilled version of the CPRetriever-Prob-Qwen3-4B-2510 model, obtained through representation alignment distillation on the Qwen3-embedding-0.6B backbone.
Key Features
- Efficient: 0.6B parameters with comparable performance to the 4B teacher model
- Alignment: Embeddings are aligned with the teacher model and can be used interchangeably
- Bilingual: Optimized for Chinese and English programming problem retrieval
Usage
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained("abcba909/CPRetriever-qwen-0.6B-distilled", trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained("abcba909/CPRetriever-qwen-0.6B-distilled")
texts = ["求N以内所有素数", "动态规划入门题目", "计算两个数的最大公约数"]
inputs = tokenizer(texts, padding=True, return_tensors="pt")
embeddings = model(**inputs)
Original Teacher Model
For detailed information about the original model architecture, training pipeline, and evaluation benchmarks, please refer to the teacher model: coldchair16/CPRetriever-Prob-Qwen3-4B-2510
- Type
- cp_embedding
- License
- cc-by-nc-4.0
- Created
- Feb 7, 2026
- Updated
- Mar 26, 2026