section3_tf.ipynb

Kernel: Unknown Kernel

Tinh chỉnh một mô hình ngôn ngữ bị ẩn đi (TensorFlow)

Install the Transformers, Datasets, and Evaluate libraries to run this notebook.

!pip install datasets evaluate transformers[sentencepiece]
!apt install git-lfs

You will need to setup git, adapt your email and name in the following cell.

!git config --global user.email "[email protected]"
!git config --global user.name "Your Name"

You will also need to be logged in to the Hugging Face Hub. Execute the following and enter your credentials.

from huggingface_hub import notebook_login

notebook_login()

from transformers import TFAutoModelForMaskedLM

model_checkpoint = "distilbert-base-uncased"
model = TFAutoModelForMaskedLM.from_pretrained(model_checkpoint)

model(model.dummy_inputs)  # Xây dựng mô hình
model.summary()

Model: "tf_distil_bert_for_masked_lm"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
distilbert (TFDistilBertMain multiple                  66362880  
_________________________________________________________________
vocab_transform (Dense)      multiple                  590592    
_________________________________________________________________
vocab_layer_norm (LayerNorma multiple                  1536      
_________________________________________________________________
vocab_projector (TFDistilBer multiple                  23866170  
=================================================================
Total params: 66,985,530
Trainable params: 66,985,530
Non-trainable params: 0
_________________________________________________________________

text = "This is a great [MASK]."

from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained(model_checkpoint)

import numpy as np
import tensorflow as tf

inputs = tokenizer(text, return_tensors="np")
token_logits = model(**inputs).logits
# Tìm vị trí [MASK] và trích xuất logit
mask_token_index = np.argwhere(inputs["input_ids"] == tokenizer.mask_token_id)[0, 1]
mask_token_logits = token_logits[0, mask_token_index, :]
# PChọn ứng viên cho [MASK] với logit cao nhất
# Chúng ta phủ định mảng trước argsort để lấy logits lớn nhất chứ không phải nhỏ nhất
top_5_tokens = np.argsort(-mask_token_logits)[:5].tolist()

for token in top_5_tokens:
    print(f">>> {text.replace(tokenizer.mask_token, tokenizer.decode([token]))}")

'>>> This is a great deal.'
'>>> This is a great success.'
'>>> This is a great adventure.'
'>>> This is a great idea.'
'>>> This is a great feat.'

from datasets import load_dataset

imdb_dataset = load_dataset("imdb")
imdb_dataset

DatasetDict({
    train: Dataset({
        features: ['text', 'label'],
        num_rows: 25000
    })
    test: Dataset({
        features: ['text', 'label'],
        num_rows: 25000
    })
    unsupervised: Dataset({
        features: ['text', 'label'],
        num_rows: 50000
    })
})

sample = imdb_dataset["train"].shuffle(seed=42).select(range(3))

for row in sample:
    print(f"\n'>>> Review: {row['text']}'")
    print(f"'>>> Label: {row['label']}'")

'>>> Review: This is your typical Priyadarshan movie--a bunch of loony characters out on some silly mission. His signature climax has the entire cast of the film coming together and fighting each other in some crazy moshpit over hidden money. Whether it is a winning lottery ticket in Malamaal Weekly, black money in Hera Pheri, "kodokoo" in Phir Hera Pheri, etc., etc., the director is becoming ridiculously predictable. Don\'t get me wrong; as clichéd and preposterous his movies may be, I usually end up enjoying the comedy. However, in most his previous movies there has actually been some good humor, (Hungama and Hera Pheri being noteworthy ones). Now, the hilarity of his films is fading as he is using the same formula over and over again.<br /><br />Songs are good. Tanushree Datta looks awesome. Rajpal Yadav is irritating, and Tusshar is not a whole lot better. Kunal Khemu is OK, and Sharman Joshi is the best.'
'>>> Label: 0'

'>>> Review: Okay, the story makes no sense, the characters lack any dimensionally, the best dialogue is ad-libs about the low quality of movie, the cinematography is dismal, and only editing saves a bit of the muddle, but Sam" Peckinpah directed the film. Somehow, his direction is not enough. For those who appreciate Peckinpah and his great work, this movie is a disappointment. Even a great cast cannot redeem the time the viewer wastes with this minimal effort.<br /><br />The proper response to the movie is the contempt that the director San Peckinpah, James Caan, Robert Duvall, Burt Young, Bo Hopkins, Arthur Hill, and even Gig Young bring to their work. Watch the great Peckinpah films. Skip this mess.'
'>>> Label: 0'

'>>> Review: I saw this movie at the theaters when I was about 6 or 7 years old. I loved it then, and have recently come to own a VHS version. <br /><br />My 4 and 6 year old children love this movie and have been asking again and again to watch it. <br /><br />I have enjoyed watching it again too. Though I have to admit it is not as good on a little TV.<br /><br />I do not have older children so I do not know what they would think of it. <br /><br />The songs are very cute. My daughter keeps singing them over and over.<br /><br />Hope this helps.'
'>>> Label: 1'

def tokenize_function(examples):
    result = tokenizer(examples["text"])
    if tokenizer.is_fast:
        result["word_ids"] = [result.word_ids(i) for i in range(len(result["input_ids"]))]
    return result


# Dùng batched=True để kích hoạt đa luồng nhanh!
tokenized_datasets = imdb_dataset.map(
    tokenize_function, batched=True, remove_columns=["text", "label"]
)
tokenized_datasets

DatasetDict({
    train: Dataset({
        features: ['attention_mask', 'input_ids', 'word_ids'],
        num_rows: 25000
    })
    test: Dataset({
        features: ['attention_mask', 'input_ids', 'word_ids'],
        num_rows: 25000
    })
    unsupervised: Dataset({
        features: ['attention_mask', 'input_ids', 'word_ids'],
        num_rows: 50000
    })
})

tokenizer.model_max_length

512

chunk_size = 128

# Tạo ra một danh sách các danh sách cho từng đặc trưng
tokenized_samples = tokenized_datasets["train"][:3]

for idx, sample in enumerate(tokenized_samples["input_ids"]):
    print(f"'>>> Review {idx} length: {len(sample)}'")

'>>> Review 0 length: 200'
'>>> Review 1 length: 559'
'>>> Review 2 length: 192'

concatenated_examples = {
    k: sum(tokenized_samples[k], []) for k in tokenized_samples.keys()
}
total_length = len(concatenated_examples["input_ids"])
print(f"'>>> Concatenated reviews length: {total_length}'")

'>>> Concatenated reviews length: 951'

chunks = {
    k: [t[i : i + chunk_size] for i in range(0, total_length, chunk_size)]
    for k, t in concatenated_examples.items()
}

for chunk in chunks["input_ids"]:
    print(f"'>>> Chunk length: {len(chunk)}'")

'>>> Chunk length: 128'
'>>> Chunk length: 128'
'>>> Chunk length: 128'
'>>> Chunk length: 128'
'>>> Chunk length: 128'
'>>> Chunk length: 128'
'>>> Chunk length: 128'
'>>> Chunk length: 55'

def group_texts(examples):
    # Nối tất cả các văn bản
    concatenated_examples = {k: sum(examples[k], []) for k in examples.keys()}
    # Tính độ dài của các văn bản được nối
    total_length = len(concatenated_examples[list(examples.keys())[0]])
    # Chúng tôi bỏ đoạn cuối cùng nếu nó nhỏ hơn chunk_size
    total_length = (total_length // chunk_size) * chunk_size
    # Chia phần theo max_len
    result = {
        k: [t[i : i + chunk_size] for i in range(0, total_length, chunk_size)]
        for k, t in concatenated_examples.items()
    }
    # Tạo cột nhãn mới
    result["labels"] = result["input_ids"].copy()
    return result

lm_datasets = tokenized_datasets.map(group_texts, batched=True)
lm_datasets

DatasetDict({
    train: Dataset({
        features: ['attention_mask', 'input_ids', 'labels', 'word_ids'],
        num_rows: 61289
    })
    test: Dataset({
        features: ['attention_mask', 'input_ids', 'labels', 'word_ids'],
        num_rows: 59905
    })
    unsupervised: Dataset({
        features: ['attention_mask', 'input_ids', 'labels', 'word_ids'],
        num_rows: 122963
    })
})

tokenizer.decode(lm_datasets["train"][1]["input_ids"])

".... at.......... high. a classic line : inspector : i'm here to sack one of your teachers. student : welcome to bromwell high. i expect that many adults of my age think that bromwell high is far fetched. what a pity that it isn't! [SEP] [CLS] homelessness ( or houselessness as george carlin stated ) has been an issue for years but never a plan to help those on the street that were once considered human who did everything from going to school, work, or vote for the matter. most people think of the homeless"

from transformers import DataCollatorForLanguageModeling

data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm_probability=0.15)

samples = [lm_datasets["train"][i] for i in range(2)]
for sample in samples:
    _ = sample.pop("word_ids")

for chunk in data_collator(samples)["input_ids"]:
    print(f"\n'>>> {tokenizer.decode(chunk)}'")

import collections
import numpy as np

from transformers.data.data_collator import tf_default_data_collator

wwm_probability = 0.2


def whole_word_masking_data_collator(features):
    for feature in features:
        word_ids = feature.pop("word_ids")

        # Tạo ra ánh xạ giữa các từ và chỉ mục token tương ứng
        mapping = collections.defaultdict(list)
        current_word_index = -1
        current_word = None
        for idx, word_id in enumerate(word_ids):
            if word_id is not None:
                if word_id != current_word:
                    current_word = word_id
                    current_word_index += 1
                mapping[current_word_index].append(idx)

        # Che ngẫu nhiền từ
        mask = np.random.binomial(1, wwm_probability, (len(mapping),))
        input_ids = feature["input_ids"]
        labels = feature["labels"]
        new_labels = [-100] * len(labels)
        for word_id in np.where(mask)[0]:
            word_id = word_id.item()
            for idx in mapping[word_id]:
                new_labels[idx] = labels[idx]
                input_ids[idx] = tokenizer.mask_token_id
        feature["labels"] = new_labels

    return tf_default_data_collator(features)

samples = [lm_datasets["train"][i] for i in range(2)]
batch = whole_word_masking_data_collator(samples)

for chunk in batch["input_ids"]:
    print(f"\n'>>> {tokenizer.decode(chunk)}'")

'>>> [CLS] bromwell high is a cartoon comedy [MASK] it ran at the same time as some other programs about school life, such as " teachers ". my 35 years in the teaching profession lead me to believe that bromwell high\'s satire is much closer to reality than is " teachers ". the scramble to survive financially, the insightful students who can see right through their pathetic teachers\'pomp, the pettiness of the whole situation, all remind me of the schools i knew and their students. when i saw the episode in which a student repeatedly tried to burn down the school, i immediately recalled.....'

'>>> .... [MASK] [MASK] [MASK] [MASK]....... high. a classic line : inspector : i\'m here to sack one of your teachers. student : welcome to bromwell high. i expect that many adults of my age think that bromwell high is far fetched. what a pity that it isn\'t! [SEP] [CLS] homelessness ( or houselessness as george carlin stated ) has been an issue for years but never a plan to help those on the street that were once considered human who did everything from going to school, work, or vote for the matter. most people think of the homeless'

train_size = 10_000
test_size = int(0.1 * train_size)

downsampled_dataset = lm_datasets["train"].train_test_split(
    train_size=train_size, test_size=test_size, seed=42
)
downsampled_dataset

DatasetDict({
    train: Dataset({
        features: ['attention_mask', 'input_ids', 'labels', 'word_ids'],
        num_rows: 10000
    })
    test: Dataset({
        features: ['attention_mask', 'input_ids', 'labels', 'word_ids'],
        num_rows: 1000
    })
})

from huggingface_hub import notebook_login

notebook_login()

tf_train_dataset = downsampled_dataset["train"].to_tf_dataset(
    columns=["input_ids", "attention_mask", "labels"],
    collate_fn=data_collator,
    shuffle=True,
    batch_size=32,
)

tf_eval_dataset = downsampled_dataset["test"].to_tf_dataset(
    columns=["input_ids", "attention_mask", "labels"],
    collate_fn=data_collator,
    shuffle=False,
    batch_size=32,
)

from transformers import create_optimizer
from transformers.keras_callbacks import PushToHubCallback
import tensorflow as tf

num_train_steps = len(tf_train_dataset)
optimizer, schedule = create_optimizer(
    init_lr=2e-5,
    num_warmup_steps=1_000,
    num_train_steps=num_train_steps,
    weight_decay_rate=0.01,
)
model.compile(optimizer=optimizer)

# Huấn luyện trong mixed-precision float16
tf.keras.mixed_precision.set_global_policy("mixed_float16")

callback = PushToHubCallback(
    output_dir=f"{model_name}-finetuned-imdb", tokenizer=tokenizer
)

import math

eval_loss = model.evaluate(tf_eval_dataset)
print(f"Perplexity: {math.exp(eval_loss):.2f}")

>>> Perplexity: 21.75

model.fit(tf_train_dataset, validation_data=tf_eval_dataset, callbacks=[callback])

eval_loss = model.evaluate(tf_eval_dataset)
print(f"Perplexity: {math.exp(eval_loss):.2f}")

>>> Perplexity: 11.32

from transformers import pipeline

mask_filler = pipeline(
    "fill-mask", model="huggingface-course/distilbert-base-uncased-finetuned-imdb"
)

preds = mask_filler(text)

for pred in preds:
    print(f">>> {pred['sequence']}")

'>>> this is a great movie.'
'>>> this is a great film.'
'>>> this is a great story.'
'>>> this is a great movies.'
'>>> this is a great character.'