metadata_transformer.py

from transformers import pipeline
from transformers import AutoTokenizer
import torch


model = "meta-llama/Llama-2-7b-chat-hf"


tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

def translate(schema_input, schema_target, pipeline):
    sequences = pipeline(
        '{} \n Translate the schema metadata file above to the schema: {}'.format(schema_input, schema_target),
        do_sample=True,
        top_k=10,
        num_return_sequences=1,
        eos_token_id=tokenizer.eos_token_id,
        max_length=200,
    )
    return sequences[0]['generated_text']