Hugging Face 的 Transformers 库快速入门（五）：快速分词器

from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
example = "Hello world!"
encoding = tokenizer(example)
print(type(encoding))
print('tokenizer.is_fast:', tokenizer.is_fast)
print('encoding.is_fast:', encoding.is_fast)

from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
example = "My name is Sylvain and I work at Hugging Face in Brooklyn."
encoding = tokenizer(example)
print(encoding.tokens())

print(encoding.word_ids())

from transformers import pipeline

token_classifier = pipeline("token-classification")
results = token_classifier("My name is Sylvain and I work at Hugging Face in Brooklyn.")
print(results)

from transformers import pipeline

token_classifier = pipeline("token-classification", grouped_entities=True)
results = token_classifier("My name is Sylvain and I work at Hugging Face in Brooklyn.")
print(results)

from transformers import pipeline

token_classifier = pipeline("token-classification", aggregation_strategy="max")
results = token_classifier("My name is Sylvain and I work at Hugging Face in Brooklyn.")
print(results)

from transformers import AutoTokenizer, AutoModelForTokenClassification

model_checkpoint = "dbmdz/bert-large-cased-finetuned-conll03-english"
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
model = AutoModelForTokenClassification.from_pretrained(model_checkpoint)

example = "My name is Sylvain and I work at Hugging Face in Brooklyn."
inputs = tokenizer(example, return_tensors="pt")
outputs = model(**inputs)

print(inputs["input_ids"].shape)
print(outputs.logits.shape)

print(model.config.id2label)

from transformers import AutoTokenizer, AutoModelForTokenClassification

model_checkpoint = "dbmdz/bert-large-cased-finetuned-conll03-english"
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
model = AutoModelForTokenClassification.from_pretrained(model_checkpoint)

example = "My name is Sylvain and I work at Hugging Face in Brooklyn."
inputs = tokenizer(example, return_tensors="pt")
outputs = model(**inputs)

import torch

probabilities = torch.nn.functional.softmax(outputs.logits, dim=-1)[0].tolist()
predictions = outputs.logits.argmax(dim=-1)[0].tolist()
print(predictions)

results = []
tokens = inputs.tokens()

for idx, pred in enumerate(predictions):
    label = model.config.id2label[pred]
    if label != "O":
        results.append(
            {"entity": label, "score": probabilities[idx][pred], "word": tokens[idx]}
        )

print(results)

inputs_with_offsets = tokenizer(example, return_offsets_mapping=True)
offset_mapping = inputs_with_offsets["offset_mapping"]
print(offset_mapping)

from transformers import AutoTokenizer, AutoModelForTokenClassification

model_checkpoint = "dbmdz/bert-large-cased-finetuned-conll03-english"
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
model = AutoModelForTokenClassification.from_pretrained(model_checkpoint)

example = "My name is Sylvain and I work at Hugging Face in Brooklyn."
inputs = tokenizer(example, return_tensors="pt")
outputs = model(**inputs)

import torch

probabilities = torch.nn.functional.softmax(outputs.logits, dim=-1)[0].tolist()
predictions = outputs.logits.argmax(dim=-1)[0].tolist()

results = []
inputs_with_offsets = tokenizer(example, return_offsets_mapping=True)
tokens = inputs_with_offsets.tokens()
offsets = inputs_with_offsets["offset_mapping"]

for idx, pred in enumerate(predictions):
    label = model.config.id2label[pred]
    if label != "O":
        start, end = offsets[idx]
        results.append(
            {
                "entity": label,
                "score": probabilities[idx][pred],
                "word": tokens[idx],
                "start": start,
                "end": end,
            }
        )

print(results)

from transformers import AutoTokenizer, AutoModelForTokenClassification

model_checkpoint = "dbmdz/bert-large-cased-finetuned-conll03-english"
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
model = AutoModelForTokenClassification.from_pretrained(model_checkpoint)

example = "My name is Sylvain and I work at Hugging Face in Brooklyn."
inputs = tokenizer(example, return_tensors="pt")
outputs = model(**inputs)

import torch

probabilities = torch.nn.functional.softmax(outputs.logits, dim=-1)[0].tolist()
predictions = outputs.logits.argmax(dim=-1)[0].tolist()

import numpy as np

results = []
inputs_with_offsets = tokenizer(example, return_offsets_mapping=True)
tokens = inputs_with_offsets.tokens()
offsets = inputs_with_offsets["offset_mapping"]

idx = 0
while idx < len(predictions):
    pred = predictions[idx]
    label = model.config.id2label[pred]
    if label != "O":
        label = label[2:] # Remove the B- or I-
        start, end = offsets[idx]
        all_scores = [probabilities[idx][pred]]
        # Grab all the tokens labeled with I-label
        while (
            idx + 1 < len(predictions)
            and model.config.id2label[predictions[idx + 1]] == f"I-{label}"
        ):
            all_scores.append(probabilities[idx + 1][predictions[idx + 1]])
            _, end = offsets[idx + 1]
            idx += 1

        score = np.mean(all_scores).item()
        word = example[start:end]
        results.append(
            {
                "entity_group": label,
                "score": score,
                "word": word,
                "start": start,
                "end": end,
            }
        )
    idx += 1

print(results)

from transformers import pipeline

question_answerer = pipeline("question-answering")
context = """
Transformers is backed by the three most popular deep learning libraries — Jax, PyTorch, and TensorFlow — with a seamless integration between them. It's straightforward to train your models with one before loading them for inference with the other.
"""
question = "Which deep learning libraries back Transformers?"
results = question_answerer(question=question, context=context)
print(results)

from transformers import AutoTokenizer, AutoModelForQuestionAnswering

model_checkpoint = "distilbert-base-cased-distilled-squad"
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
model = AutoModelForQuestionAnswering.from_pretrained(model_checkpoint)

context = """
Transformers is backed by the three most popular deep learning libraries — Jax, PyTorch, and TensorFlow — with a seamless integration between them. It's straightforward to train your models with one before loading them for inference with the other.
"""
question = "Which deep learning libraries back Transformers?"

inputs = tokenizer(question, context, return_tensors="pt")
outputs = model(**inputs)

print(inputs["input_ids"].shape)

start_logits = outputs.start_logits
end_logits = outputs.end_logits
print(start_logits.shape, end_logits.shape)

from transformers import AutoTokenizer, AutoModelForQuestionAnswering

model_checkpoint = "distilbert-base-cased-distilled-squad"
tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)
model = AutoModelForQuestionAnswering.from_pretrained(model_checkpoint)

context = """
Transformers is backed by the three most popular deep learning libraries — Jax, PyTorch, and TensorFlow — with a seamless integration between them. It's straightforward to train your models with one before loading them for inference with the other.
"""
question = "Which deep learning libraries back Transformers?"

inputs = tokenizer(question, context, return_tensors="pt")
outputs = model(**inputs)

start_logits = outputs.start_logits
end_logits = outputs.end_logits

import torch

sequence_ids = inputs.sequence_ids()
mask = [i != 1 for i in sequence_ids]
mask[0] = False # Unmask the [CLS] token
mask = torch.tensor(mask)[None]

start_logits[mask] = -10000
end_logits[mask] = -10000

start_probabilities = torch.nn.functional.softmax(start_logits, dim=-1)[0]
end_probabilities = torch.nn.functional.softmax(end_logits, dim=-1)[0]

scores = start_probabilities[:, None] * end_probabilities[None, :]
scores = torch.triu(scores)

max_index = scores.argmax().item()
start_index = max_index // scores.shape[1]
end_index = max_index % scores.shape[1]

inputs_with_offsets = tokenizer(question, context, return_offsets_mapping=True)
offsets = inputs_with_offsets["offset_mapping"]

start, _ = offsets[start_index]
_, end = offsets[end_index]

result = {
    "answer": context[start:end],
    "start": start,
    "end": end,
    "score": float(scores[start_index, end_index]),
}
print(result)

inputs = tokenizer(question, long_context, max_length=384, truncation="only_second")

from transformers import pipeline

question_answerer = pipeline("question-answering")

long_context = """
Transformers: State of the Art NLP

Transformers provides thousands of pretrained models to perform tasks on texts such as classification, information extraction, question answering, summarization, translation, text generation and more in over 100 languages. Its aim is to make cutting-edge NLP easier to use for everyone.

Transformers provides APIs to quickly download and use those pretrained models on a given text, fine-tune them on your own datasets and then share them with the community on our model hub. At the same time, each python module defining an architecture is fully standalone and can be modified to enable quick research experiments.

Why should I use transformers?

1. Easy-to-use state-of-the-art models:
  - High performance on NLU and NLG tasks.
  - Low barrier to entry for educators and practitioners.
  - Few user-facing abstractions with just three classes to learn.
  - A unified API for using all our pretrained models.
  - Lower compute costs, smaller carbon footprint:

2. Researchers can share trained models instead of always retraining.
  - Practitioners can reduce compute time and production costs.
  - Dozens of architectures with over 10,000 pretrained models, some in more than 100 languages.

3. Choose the right framework for every part of a model's lifetime:
  - Train state-of-the-art models in 3 lines of code.
  - Move a single model between TF2.0/PyTorch frameworks at will.
  - Seamlessly pick the right framework for training, evaluation and production.

4. Easily customize a model or an example to your needs:
  - We provide examples for each architecture to reproduce the results published by its original authors.
  - Model internals are exposed as consistently as possible.
  - Model files can be used independently of the library for quick experiments.

Transformers is backed by the three most popular deep learning libraries — Jax, PyTorch and TensorFlow — with a seamless integration between them. It's straightforward to train your models with one before loading them for inference with the other.
"""
question = "Which deep learning libraries back Transformers?"
results = question_answerer(question=question, context=long_context)
print(results)

sentence = "This sentence is not too long but we are going to split it anyway."
inputs = tokenizer(
    sentence, truncation=True, return_overflowing_tokens=True, max_length=6, stride=2
)

for ids in inputs["input_ids"]:
    print(tokenizer.decode(ids))

sentence = "This sentence is not too long but we are going to split it anyway."
inputs = tokenizer(
    sentence, truncation=True, return_overflowing_tokens=True, max_length=6, stride=2
)
print(inputs.keys())
print(inputs["overflow_to_sample_mapping"])

sentences = [
    "This sentence is not too long but we are going to split it anyway.",
    "This sentence is shorter but will still get split.",
]
inputs = tokenizer(
    sentences, truncation=True, return_overflowing_tokens=True, max_length=6, stride=2
)
print(inputs["overflow_to_sample_mapping"])

inputs = tokenizer(
    question,
    long_context,
    stride=128,
    max_length=384,
    padding="longest",
    truncation="only_second",
    return_overflowing_tokens=True,
    return_offsets_mapping=True,
)

_ = inputs.pop("overflow_to_sample_mapping")
offsets = inputs.pop("offset_mapping")

inputs = inputs.convert_to_tensors("pt")
print(inputs["input_ids"].shape)

outputs = model(**inputs)

start_logits = outputs.start_logits
end_logits = outputs.end_logits
print(start_logits.shape, end_logits.shape)

sequence_ids = inputs.sequence_ids()
# Mask everything apart from the tokens of the context
mask = [i != 1 for i in sequence_ids]
# Unmask the [CLS] token
mask[0] = False
# Mask all the [PAD] tokens
mask = torch.logical_or(torch.tensor(mask)[None], (inputs["attention_mask"] == 0))

start_logits[mask] = -10000
end_logits[mask] = -10000

start_probabilities = torch.nn.functional.softmax(start_logits, dim=-1)
end_probabilities = torch.nn.functional.softmax(end_logits, dim=-1)

candidates = []
for start_probs, end_probs in zip(start_probabilities, end_probabilities):
    scores = start_probs[:, None] * end_probs[None, :]
    idx = torch.triu(scores).argmax().item()

    start_idx = idx // scores.shape[0]
    end_idx = idx % scores.shape[0]
    score = scores[start_idx, end_idx].item()
    candidates.append((start_idx, end_idx, score))

for candidate, offset in zip(candidates, offsets):
    start_token, end_token, score = candidate
    start_char, _ = offset[start_token]
    _, end_char = offset[end_token]
    answer = long_context[start_char:end_char]
    result = {"answer": answer, "start": start_char, "end": end_char, "score": score}
    print(result)