metadata

language:
  - hu
  - en
  - zh
tags:
  - text-generation
license: cc-by-nc-4.0
widget:
  - text: Elmesélek egy történetet a nyelvtechnológiáról.

PULI GPTrio (6.7 billion parameter)

For further details, see our demo site.

Hungarian-English-Chinese trilingual GPT-NeoX model (6.7 billion parameter)
Trained with EleutherAI's GPT-NeoX github
Checkpoint: 410 000 steps

Dataset

Hungarian: 41 508 933 801 words (314 GB)
English: 61 906 491 82 words (391 GB)
Github: 6 018 366 documents (33 GB)
Chinese: 98 693 705 456 Chinese character (340 GB)
- (12 072 234 774 non Chinese token)

Limitations

max_seq_length = 2048
float16

Citation

If you use this model, please cite the following paper:

@inproceedings {yang-puli-gptrio,
    title = {Mono- and multilingual GPT-3 models for Hungarian},
    booktitle = {Text, Speech, and Dialogue - 26th International Conference, {TSD} 2023, Proceedings},
    year = {2023},
    publisher = {Springer},
    series = {Lecture Notes in Computer Science},
    address = {Plzeň, Czech Republic},
    author = {Yang, Zijian Győző and Laki, László János and Váradi, Tamás and Prószéky, Gábor},
    pages = {Accepted}
}

Usage

from transformers import GPTNeoXForCausalLM, GPTNeoXTokenizerFast

model = GPTNeoXForCausalLM.from_pretrained("NYTK/PULI-GPTrio")
tokenizer = GPTNeoXTokenizerFast.from_pretrained("NYTK/PULI-GPTrio")
prompt = "Elmesélek egy történetet a nyelvtechnológiáról."
input_ids = tokenizer(prompt, return_tensors="pt").input_ids

gen_tokens = model.generate(
    input_ids,
    do_sample=True,
    temperature=0.9,
    max_length=100,
)

gen_text = tokenizer.batch_decode(gen_tokens)[0]
print(gen_text)

Usage with pipeline

from transformers import pipeline, GPTNeoXForCausalLM, GPTNeoXTokenizerFast

model = GPTNeoXForCausalLM.from_pretrained("NYTK/PULI-GPTrio")
tokenizer = GPTNeoXTokenizerFast.from_pretrained("NYTK/PULI-GPTrio")
prompt = "Elmesélek egy történetet a nyelvtechnológiáról."
generator = pipeline(task="text-generation", model=model, tokenizer=tokenizer)

print(generator(prompt)[0]["generated_text"])