Edit model card

Mengzi-GPT-neo model (Chinese)

Pretrained model on 300G Chinese corpus.

Usage

import torch
import sentencepiece as spm
from transformers import GPTNeoForCausalLM
tokenizer = spm.SentencePieceProcessor(model_file="mengzi_gpt.model")
model = GPTNeoForCausalLM.from_pretrained("Langboat/mengzi-gpt-neo-base")

def lm(prompt, top_k, top_p, max_length, repetition_penalty):
    input_ids = torch.tensor(tokenizer.encode([prompt]), dtype=torch.long, device='cuda')
    gen_tokens = model.generate(
        input_ids,
        do_sample=True,
        top_k=top_k,
        top_p=top_p,
        max_length=max_length+len(prompt),
        repetition_penalty=repetition_penalty)
    result = tokenizer.decode(gen_tokens.tolist())[0]
    return result
Downloads last month
57
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.

Spaces using Langboat/mengzi-gpt-neo-base 2