Create README.md
Browse files主要是从部分opus100和sahil2801/CodeAlpaca-20k中的英文提示词作为翻译原文，用chatglm作为翻译器翻译成中文作为数据，其中原始模型为Helsinki-NLP/opus-mt-en-zh,利用Seq2SeqTrainer做微调。
使用方法：
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer, pipeline
parser.add_argument('--device', default="cpu", type=str, help='"cuda:1"、"cuda:2"……')
mode_name = "DDDSSS/translation_en-zh"
model = AutoModelForSeq2SeqLM.from_pretrained(mode_name)
tokenizer = AutoTokenizer.from_pretrained(mode_name)
translation = pipeline("translation_en_to_zh", model=model, tokenizer=tokenizer,
torch_dtype="float", device_map=True,device=device)
x = ["You can even specify your model’s eval results in a structured way, which will allow the Hub to parse, display, and even link them to Papers With Code leaderboards. ","If nothing is detected and there is a config.json file, it’s assumed the library is transformers."]
response = translation(x, max_length=450)
print("翻译为：",response)
微调方法：
from datasets import load_dataset, load_from_disk
from transformers import AutoTokenizer
from transformers import DataCollatorForSeq2Seq
import evaluate
import numpy as np
from transformers import AutoModelForSeq2SeqLM, Seq2SeqTrainingArguments, Seq2SeqTrainer
books = load_dataset("json", data_files=".json")
books = books["train"].train_test_split(test_size=0.2)
checkpoint = "./opus-mt-en-zh"
tokenizer = AutoTokenizer.from_pretrained(checkpoint)
source_lang = "en"
target_lang = "zh"
def preprocess_function(examples):
inputs = [example[source_lang] for example in examples["translation"]]
targets = [example[target_lang] for example in examples["translation"]]
model_inputs = tokenizer(inputs, text_target=targets, max_length=512, truncation=True)
return model_inputs
tokenized_books = books.map(preprocess_function, batched=True)
data_collator = DataCollatorForSeq2Seq(tokenizer=tokenizer, model=checkpoint)
metric = evaluate.load("sacrebleu")
def postprocess_text(preds, labels):
preds = [pred.strip() for pred in preds]
labels = [[label.strip()] for label in labels]
return preds, labels
def compute_metrics(eval_preds):
preds, labels = eval_preds
if isinstance(preds, tuple):
preds = preds[0]
decoded_preds = tokenizer.batch_decode(preds, skip_special_tokens=True)
labels = np.where(labels != -100, labels, tokenizer.pad_token_id)
decoded_labels = tokenizer.batch_decode(labels, skip_special_tokens=True)
decoded_preds, decoded_labels = postprocess_text(decoded_preds, decoded_labels)
result = metric.compute(predictions=decoded_preds, references=decoded_labels)
result = {"bleu": result["score"]}
prediction_lens = [np.count_nonzero(pred != tokenizer.pad_token_id) for pred in preds]
result["gen_len"] = np.mean(prediction_lens)
result = {k: round(v, 4) for k, v in result.items()}
return result
model = AutoModelForSeq2SeqLM.from_pretrained(checkpoint)
batchsize=4
training_args = Seq2SeqTrainingArguments(
output_dir="./my_awesome_opus_books_model",
evaluation_strategy="epoch",
learning_rate=2e-4,
per_device_train_batch_size=batchsize,
per_device_eval_batch_size=batchsize,
weight_decay=0.01,
# save_total_limit=3,
num_train_epochs=4,
predict_with_generate=True,
fp16=True,
push_to_hub=False,
save_strategy="epoch",
jit_mode_eval=True
)
trainer = Seq2SeqTrainer(
model=model,
args=training_args,
train_dataset=tokenized_books["train"],
eval_dataset=tokenized_books["test"],
tokenizer=tokenizer,
data_collator=data_collator,
compute_metrics=compute_metrics,
)
trainer.train()