File size: 1,793 Bytes
d870346 16014b0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 |
---
license: mit
---
language:
- en
pipeline_tag: text-generation
tags:
---
# Qwen2-1.5B-Sign
## Introduction
Qwen2-Sign is a text to sige model base on Qwen2.
## Finetune Details
- Finetune dataset: [alpaca-zh-text2sign](https://huggingface.co/datasets/thundax/alpaca-zh-text2sign)
- Finetune parameter
| Parameter | Value |
|------------------------------|--------|
| learning_rate | 5e-05 |
| train_batch_size | 4 |
| eval_batch_size | 4 |
| gradient_accumulation_steps | 8 |
| total_train_batch_size | 32 |
| lr_scheduler_type | cosine |
| lr_scheduler_warmup_steps | 100 |
| num_epochs | 4 |
## Quickstart
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
device = "cuda" # the device to load the model onto
model = AutoModelForCausalLM.from_pretrained(
"thundax/Qwen2-1.5B-Sign",
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("thundax/Qwen2-1.5B-Sign")
text = "你好，世界！"
text = f'Translate sentence into labels\n{text}\n'
model_inputs = tokenizer([text], return_tensors="pt").to(device)
generated_ids = model.generate(
model_inputs.input_ids,
max_new_tokens=512
)
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]
```
## Citation
If you find our work helpful, feel free to give us a cite.
```
@software{qwen2-sign,
author = {thundax},
title = {qwen2-sign: A Tool for Text to Sign},
year = {2024},
url = {https://github.com/thundax-lyp},
}
```
|