This model is based on bigscience/bloom-1b7.

We pruned its vocabulary from 250880 to 46145 with Chinese corpus to reduce GPU memory usage. So the total parameter is 1.4b now.

How to use

from transformers import BloomTokenizerFast, BloomForCausalLM

tokenizer = BloomTokenizerFast.from_pretrained('Langboat/bloom-1b4-zh')
model = BloomForCausalLM.from_pretrained('Langboat/bloom-1b4-zh')

print(tokenizer.batch_decode(model.generate(tokenizer.encode('中国的首都是', return_tensors='pt'))))

Downloads last month: 6,629

Safetensors

Model size

1B params

Tensor type

F32

Langboat
/

bloom-1b4-zh

How to use

Spaces using Langboat/bloom-1b4-zh 11