# **NanoTranslator-S**

[English](README.md) | 简体中文

## Introduction

这是 NanoTranslator 的 Small 型号，目前仅支持**英译中**。仓库中同时提供了 ONNX 版本的模型。


| Size | Params. |  V.  |  H.  |  I.  |  L.  | Att. H. | KV H. | Tie Emb. |
| :--: | :-----: | :--: | :--: | :--: | :--: | :-----: | :---: | :------: |
|  XL  |  50 M   | 8000 | 320  | 1792 |  24  |   16    |   4   |   True   |
|  L   |  22 M   | 8000 | 256  | 1408 |  16  |   16    |   4   |   True   |
|  M   |  9 M  | 4000 | 168 | 896 |  16  |   12    |   4   |   True   |
|  S   | 2 M  | 2000 |  96  | 512  |  12  |   12    |   4   |   True   |

- **V.** - vocab size
- **H.** - hidden size
- **I.** - intermediate size
- **L.** - num layers
- **Att. H.** - num attention heads
- **KV H.** - num kv heads
- **Tie Emb.** - tie word embeddings



## How to use

Prompt 格式如下：

```
<|im_start|> {English Text} <|endoftext|>
```

### Directly using transformers

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

model_path = 'Mxode/NanoTranslator-S'

tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(model_path)

def translate(text: str, model, **kwargs):
    generation_args = dict(
        max_new_tokens = kwargs.pop("max_new_tokens", 512),
        do_sample = kwargs.pop("do_sample", True),
        temperature = kwargs.pop("temperature", 0.55),
        top_p = kwargs.pop("top_p", 0.8),
        top_k = kwargs.pop("top_k", 40),
        **kwargs
    )

    prompt = "<|im_start|>" + text + "<|endoftext|>"
    model_inputs = tokenizer([prompt], return_tensors="pt").to(model.device)

    generated_ids = model.generate(model_inputs.input_ids, **generation_args)
    generated_ids = [
        output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
    ]

    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
    return response

text = "I love to watch my favorite TV series."

response = translate(text, model, max_new_tokens=64, do_sample=False)
print(response)
```


### ONNX

根据实际测试，使用 ONNX 模型推理会比直接使用 transformers 推理要**快 2～10 倍**。

如果希望使用 ONNX 模型，那么你需要手动切换到 [onnx 分支](https://huggingface.co/Mxode/NanoTranslator-S/tree/onnx)并从本地加载。

参考文档：

- [Export to ONNX](https://huggingface.co/docs/transformers/serialization)
- [Inference pipelines with the ONNX Runtime accelerator](https://huggingface.co/docs/optimum/main/en/onnxruntime/usage_guides/pipelines)

**Using ORTModelForCausalLM**

```python
from optimum.onnxruntime import ORTModelForCausalLM
from transformers import AutoTokenizer

model_path = "your/folder/to/onnx_model"

ort_model = ORTModelForCausalLM.from_pretrained(model_path)
tokenizer = AutoTokenizer.from_pretrained(model_path)

text = "I love to watch my favorite TV series."

response = translate(text, ort_model, max_new_tokens=64, do_sample=False)
print(response)
```

**Using pipeline**

```python
from optimum.pipelines import pipeline

model_path = "your/folder/to/onnx_model"
pipe = pipeline("text-generation", model=model_path, accelerator="ort")

text = "I love to watch my favorite TV series."

response = pipe(text, max_new_tokens=64, do_sample=False)
response
```