---
license: other
license_name: deepseek
license_link: https://github.com/deepseek-ai/DeepSeek-MoE/blob/main/LICENSE-MODEL
model-index:
- name: deepseek-moe-16b-base
  results:
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: AI2 Reasoning Challenge (25-Shot)
      type: ai2_arc
      config: ARC-Challenge
      split: test
      args:
        num_few_shot: 25
    metrics:
    - type: acc_norm
      value: 53.24
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=deepseek-ai/deepseek-moe-16b-base
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: HellaSwag (10-Shot)
      type: hellaswag
      split: validation
      args:
        num_few_shot: 10
    metrics:
    - type: acc_norm
      value: 79.77
      name: normalized accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=deepseek-ai/deepseek-moe-16b-base
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: MMLU (5-Shot)
      type: cais/mmlu
      config: all
      split: test
      args:
        num_few_shot: 5
    metrics:
    - type: acc
      value: 46.31
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=deepseek-ai/deepseek-moe-16b-base
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: TruthfulQA (0-shot)
      type: truthful_qa
      config: multiple_choice
      split: validation
      args:
        num_few_shot: 0
    metrics:
    - type: mc2
      value: 36.08
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=deepseek-ai/deepseek-moe-16b-base
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: Winogrande (5-shot)
      type: winogrande
      config: winogrande_xl
      split: validation
      args:
        num_few_shot: 5
    metrics:
    - type: acc
      value: 73.72
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=deepseek-ai/deepseek-moe-16b-base
      name: Open LLM Leaderboard
  - task:
      type: text-generation
      name: Text Generation
    dataset:
      name: GSM8k (5-shot)
      type: gsm8k
      config: main
      split: test
      args:
        num_few_shot: 5
    metrics:
    - type: acc
      value: 17.29
      name: accuracy
    source:
      url: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=deepseek-ai/deepseek-moe-16b-base
      name: Open LLM Leaderboard
---

<p align="center">
<img width="500px" alt="DeepSeek Chat" src="https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/images/logo.png?raw=true">
</p>
<p align="center"><a href="https://www.deepseek.com/">[🏠Homepage]</a>  |  <a href="https://chat.deepseek.com/">[🤖 Chat with DeepSeek LLM]</a>  |  <a href="https://discord.gg/Tc7c45Zzu5">[Discord]</a>  |  <a href="https://github.com/deepseek-ai/DeepSeek-LLM/blob/main/images/qr.jpeg">[Wechat(微信)]</a> </p>

<p align="center">
  <a href="https://arxiv.org/pdf/2401.06066.pdf"><b>Paper Link</b>👁️</a>
</p>
<hr>




### 1. Introduction to DeepSeekMoE
See the [Introduction](https://github.com/deepseek-ai/DeepSeek-MoE/blob/main) for more details.

### 2. How to Use
Here give some examples of how to use our model.
#### Text Completion
```python
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig

model_name = "deepseek-ai/deepseek-moe-16b-base"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto")
model.generation_config = GenerationConfig.from_pretrained(model_name)
model.generation_config.pad_token_id = model.generation_config.eos_token_id

text = "An attention function can be described as mapping a query and a set of key-value pairs to an output, where the query, keys, values, and output are all vectors. The output is"
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs.to(model.device), max_new_tokens=100)

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

### 3. License
This code repository is licensed under the MIT License. The use of DeepSeekMoE models is subject to the Model License. DeepSeekMoE supports commercial use.

See the [LICENSE-MODEL](https://github.com/deepseek-ai/DeepSeek-MoE/blob/main/LICENSE-MODEL) for more details.

### 4. Contact

If you have any questions, please raise an issue or contact us at [service@deepseek.com](mailto:service@deepseek.com).


# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_deepseek-ai__deepseek-moe-16b-base)

|             Metric              |Value|
|---------------------------------|----:|
|Avg.                             |51.07|
|AI2 Reasoning Challenge (25-Shot)|53.24|
|HellaSwag (10-Shot)              |79.77|
|MMLU (5-Shot)                    |46.31|
|TruthfulQA (0-shot)              |36.08|
|Winogrande (5-shot)              |73.72|
|GSM8k (5-shot)                   |17.29|