Edit model card

Introduction

The model was trained to translate a single sentence from English to Korean with a 1.3M dataset in the technology science domain. Dataset: nayohan/tech_science_translation

Loading the Model

Use the following Python code to load the model:

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "nayohan/llama3-8b-translation-en-ko-1sent"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
  model_name,
  device_map="auto",
  torch_dtype=torch.bfloat16
)

Generating Text

To generate text, use the following Python code: No support for other languages or reverse direction and styles at this time.

source="en"
target="ko"
style="written"
SYSTEM_PROMPT=f"Acts as a translator. Translate {source} sentences into {target} sentences in {style} style."
s = "The aerospace industry is a flower in the field of technology and science."
conversation = [{'role': 'system', 'content': SYSTEM_PROMPT},
                {'role': 'user', 'content': s}]

inputs = tokenizer.apply_chat_template(
  conversation,
  tokenize=True,
  add_generation_prompt=True,
  return_tensors='pt'
).to("cuda")

outputs = model.generate(inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][len(inputs[0]):]))
# Result
# INPUT: <|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nActs as a translator. Translate en sentences into ko sentences in written style.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nThe aerospace industry is a flower in the field of technology and science.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n
# OUTPUT: 항공 우주 산업은 기술과 과학 분야의 꽃이라고 할 수 있다.

## [Warning!] In multiple sentences, there is a tendency to output in a single sentence.
# INPUT: <|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nActs as a translator. Translate ko sentences into en sentences in written style.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\n
Technical and basic sciences are very important in terms of research. It has a significant impact on the industrial development of a country. Government policies control the research budget.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n
# OUTPUT: 연구 측면에서 기술 및 기초 과학은 국가의 산업 발전에 큰 영향을 미치며 정부 정책은 연구 예산을 통제한다.

Citation

@article{llama3modelcard,
        title={Llama 3 Model Card},
        author={AI@Meta},
        year={2024},
        url={https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md}
}

Our trainig code can be found here: [TBD]

Downloads last month
21
Safetensors
Model size
8.03B params
Tensor type
BF16
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for nayohan/llama3-8b-it-translation-tech-en-ko-1sent

Finetuned
(416)
this model

Dataset used to train nayohan/llama3-8b-it-translation-tech-en-ko-1sent