---
library_name: transformers
license: apache-2.0
datasets:
- HuggingFaceTB/smollm-corpus
language:
- en
pipeline_tag: text-generation
---


# **Doge 160M**

<div align="center">
  <img src="https://huggingface.co/spaces/SmallDoge/README/resolve/main/org_icon.png" width="100%" alt="SmallDoge" />
</div>
<hr>
<div align="center">
  <a href="https://arxiv.org/abs/2412.11834" target="_blank" style="margin: 2px;">
    <img alt="arXiv" src="https://img.shields.io/static/v1?label=arXiv&message=2412.11834&color=B31B1B&logo=arXiv" style="display: inline-block; vertical-align: middle;"/>
  </a>
  <a href="https://github.com/SmallDoges/small-doge" target="_blank" style="margin: 2px;">
    <img alt="GitHub" src="https://img.shields.io/badge/GitHub-SmallDoge-181717?logo=github" style="display: inline-block; vertical-align: middle;"/>
  </a>
  <a href="https://huggingface.co/SmallDoge" target="_blank" style="margin: 2px;">
    <img alt="Hugging Face" src="https://img.shields.io/badge/%F0%9F%A4%97%20HuggingFace-SmallDoge-ffc107?color=ffc107&logoColor=white" style="display: inline-block; vertical-align: middle;"/>
  </a>
  <a href="https://github.com/SmallDoges/small-doge/blob/main/LICENSE" style="margin: 2px;">
    <img alt="License" src="https://img.shields.io/badge/License-Apache--2.0-blue.svg" style="display: inline-block; vertical-align: middle;"/>
  </a>
</div>

Doge uses Dynamic Mask Attention as sequence transformation and can use Multi-Layer Perceptron or Cross Domain Mixture of Experts as state transformation. Dynamic Mask Attention allows the Transformer to use self-attention during training and state space during inference, and Cross Domain Mixture of Experts can directly inherit the weights of Multi-Layer Perceptron for further training. This model is trained by [SmallDoge](https://huggingface.co/SmallDoge) community, for detailed algorithm and model architecture, please refer to [Wonderful Matrices](https://arxiv.org/abs/2412.11834), all training details and code are publicly available on the [small-doge](https://github.com/SmallDoges/small-doge) repository.


## Uses

```python
>>> from transformers import AutoTokenizer, AutoModelForCausalLM

>>> tokenizer = AutoTokenizer.from_pretrained("SmallDoge/Doge-160M")
>>> model = AutoModelForCausalLM.from_pretrained("SmallDoge/Doge-160M", trust_remote_code=True)
>>> inputs = tokenizer("Hey how are you doing?", return_tensors="pt")

>>> out = model.generate(**inputs, max_new_tokens=100)
>>> print(tokenizer.batch_decode(out))
```


## Model Details

We build the Doge by doing Per-Training on [Smollm-Corpus](https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus).

> NOTE: If you want to continue pre-training this model, you can find the unconverged checkpoint [here](https://huggingface.co/SmallDoge/Doge-160M-checkpoint).

> NOTE: These models has not been fine-tuned for instruction, the instruction model is [here](https://huggingface.co/SmallDoge/Doge-160M-Instruct).

> TODO: The larger model is under training and will be uploaded soon.

**Pre-Training**:

| Model | Training Data | Steps | Content Length | Tokens | LR | Batch Size | Precision | RTX 4090 GPU hours |
|---|---|---|---|---|---|---|---|---|
| [Doge-20M](https://huggingface.co/SmallDoge/Doge-20M) | [HuggingFaceTB/smollm-corpus](https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus) | 8k  | 2048 | 4B | 8e-3 | 0.5M | bfloat16 | 14 |
| [Doge-60M](https://huggingface.co/SmallDoge/Doge-60M) | [HuggingFaceTB/smollm-corpus](https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus) | 16k  | 2048 | 16B | 6e-3 | 1M | bfloat16 | 128 |
| [Doge-160M](https://huggingface.co/SmallDoge/Doge-160M) | [HuggingFaceTB/smollm-corpus](https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus) | 24k  | 2048 | 32B | 4e-3 | 1.5M | bfloat16 | 522 |

**Evaluation**:

| Model | MMLU | TriviaQA | ARC | PIQA | HellaSwag | OBQA | Winogrande | tokens / s on CPU |
|---|---|---|---|---|---|---|---|---|
| [Doge-20M](https://huggingface.co/SmallDoge/Doge-20M) | 25.4 | 0.03 | 29.8 | 58.4 | 27.3 | 25.6 | 50.2 | 142 |
| [Doge-60M](https://huggingface.co/SmallDoge/Doge-60M) | 26.4 | 0.2 | 37.9 | 61.4 | 31.5 | 28.0 | 50.8 | 62 |
| [Doge-160M](https://huggingface.co/SmallDoge/Doge-160M) | 29.2 | 4.8 | 44.4 | 66.3 | 38.7 | 34.4 | 52.2 | 28 |

> All evaluations are done using five-shot settings, without additional training on the benchmarks.

**Procedure**:

[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/loser_cheems/huggingface/runs/3uyc9a89) 


**Environment**:

- Image: nvcr.io/nvidia/pytorch:24.12-py3
- Hardware: 1x NVIDIA RTX 4090
- Software: Transformers


## Citation

```bibtex
@misc{shi2024wonderfulmatrices,
      title={Wonderful Matrices: Combining for a More Efficient and Effective Foundation Model Architecture}, 
      author={Jingze Shi and Bingheng Wu},
      year={2024},
      eprint={2412.11834},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2412.11834}, 
}
```