阿葛廷

人一定要靠自己!

2 minute read

Tags: , , ,

  1. Tokenizer 文字 Input

  2. 組成 tf 看得懂的樣子

  3. Build Model

  4. Testing

Import

from transformers import BertTokenizer, TFBertForSequenceClassification
import tensorflow as tf
import pandas as pd
from tensorflow.keras.layers import Input, Dense
from tensorflow.keras.models import Model
import numpy as np

Tokenizer

# Load tokenizer
tokenizer = BertTokenizer.from_pretrained('bert-base-chinese')
# Load pre-trained model
base_model = TFBertForSequenceClassification.from_pretrained('bert-base-chinese')

Data

"""
    ("今天生產效率", 1),
    {"昨日生產效率", 1},
    ("事不是可以提高訂單量", 1),
    ("今天工廠整場表現", 1),
    ("哪幾台產出最多良品", 0),
    ("有哪些機台異常？", 0),
    ("哪些工單會遲交", 0),
    ("請問 小王 今天表現如何", 0),
    ("請問 小明 這週表現如何", 0)
    
    ----
    
        ("今天生產效率", "效率"),
    {"昨日生產效率", "效率"},
    ("事不是可以提高訂單量", "效率"),
    ("今天工廠整場表現", "效率"),
    ("哪幾台產出最多良品", "其他"),
    ("有哪些機台異常？", "其他"),
    ("哪些工單會遲交", "其他"),
    ("請問 小王 今天表現如何", "其他"),
    ("請問 小明 這週表現如何", "其他")
"""
mock_data = [    
  ("今天生產效率", 1),
    ("昨日生產效率", 1),
    ("事不是可以提高訂單量", 1),
    ("今天工廠整場表現", 1),
    ("哪幾台產出最多良品", 0),
    ("有哪些機台異常？", 0),
    ("哪些工單會遲交", 0),
    ("請問 小王 今天表現如何", 0),
    ("請問 小明 這週表現如何", 0)
]

# 0: 其他
# 1: 效率
label_array = ["其他", "效率"]
df = pd.DataFrame(mock_data, columns=['text', 'label'])

組 dataset

# Tokenize data
tokens = tokenizer.batch_encode_plus(
    df['text'].tolist(),
    max_length=512,
    padding='max_length',
    truncation=True,
    return_attention_mask=True,
    return_tensors='tf'
)

# Create TensorFlow dataset
dataset = tf.data.Dataset.from_tensor_slices(({
    'input_ids': tokens['input_ids'],
    'attention_mask': tokens['attention_mask']
}, df['label']))

Transfer Learning & Build Model

# Freeze BERT layers
for layer in base_model.layers:
    layer.trainable = False
    
# Define input layer
inputs = Input(shape=(512,), dtype=tf.int32, name="input_ids")

# Define classification layer
x = base_model(inputs)[0]
x = Dense(256, activation="relu")(x)
outputs = Dense(len(label_array), activation="softmax")(x)

model = Model(inputs=inputs, outputs=outputs)

optimizer = tf.keras.optimizers.Adam(learning_rate=2e-5, epsilon=1e-08)
model.compile(optimizer=optimizer, loss='sparse_categorical_crossentropy', metrics=['accuracy'])

# Fit model
model.fit(dataset.shuffle(100).batch(64), epochs=6, batch_size=64)

Model Summary

model.summary()


"""
Model: "model_7"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 input_ids (InputLayer)      [(None, 512)]             0         
                                                                 
 tf_bert_for_sequence_classi  TFSequenceClassifierOutp  102269186
 fication_8 (TFBertForSequen  ut(loss=None, logits=(No           
 ceClassification)           ne, 2),                             
                              hidden_states=None, att            
                             entions=None)                       
                                                                 
 dense_18 (Dense)            (None, 256)               768       
                                                                 
 dense_19 (Dense)            (None, 2)                 514       
                                                                 
=================================================================
Total params: 102,270,468
Trainable params: 1,282
Non-trainable params: 102,269,186
_________________________________________________________________
"""

Testing

questions = "今天 ggg 表現如和!"
tokens = tokenizer.encode_plus(questions,truncation=True, return_tensors='tf', max_length=512, padding='max_length')
model_input = {
    'input_ids': tokens['input_ids'],
    'attention_mask': tokens['attention_mask']
}
result = model(model_input)

Check Result

predicted_label_index = np.argmax(result)
predicted_probs = result[0]
ans_idx = np.argmax(result)
ans_pro = predicted_probs[ans_idx]
result_intent = label_array[ans_idx]

if ans_pro > 0.8:
    print("信心超過 80 %")
    print(result_intent)
else:
    print("烙賽!!! 信心只有 {}% -> {}".format(ans_pro * 100, result_intent))

"""
烙賽!!! 信心只有 50.254486083984375% -> 效率

"""

requirements.txt

absl-py==1.4.0
aiofiles==22.1.0
aiosqlite==0.18.0
anyio==3.6.2
appnope==0.1.3
argon2-cffi==21.3.0
argon2-cffi-bindings==21.2.0
arrow==1.2.3
astunparse==1.6.3
attrs==22.2.0
Babel==2.12.1
backcall==0.2.0
beautifulsoup4==4.11.2
bleach==6.0.0
cached-property==1.5.2
cachetools==5.3.0
certifi==2022.12.7
cffi==1.15.1
charset-normalizer==3.0.1
ckip-transformers==0.3.2
ckiptagger==0.2.1
click==8.1.3
debugpy==1.6.6
decorator==5.1.1
defusedxml==0.7.1
entrypoints==0.4
fastjsonschema==2.16.3
filelock==3.9.0
Flask==2.2.3
flatbuffers==23.1.21
fqdn==1.5.1
gast==0.4.0
gdown==4.6.4
google-auth==2.16.1
google-auth-oauthlib==0.4.6
google-pasta==0.2.0
grpcio==1.51.3
h5py==3.8.0
hanlp-downloader==0.0.25
huggingface-hub==0.12.1
idna==3.4
importlib-metadata==6.0.0
importlib-resources==5.12.0
ipykernel==6.16.2
ipython==7.34.0
ipython-genutils==0.2.0
isoduration==20.11.0
itsdangerous==2.1.2
jedi==0.18.2
jieba==0.42.1
Jinja2==3.1.2
JPype1==0.7.0
json5==0.9.11
jsonpointer==2.3
jsonschema==4.17.3
jupyter-events==0.6.3
jupyter-server==1.23.6
jupyter-ydoc==0.2.2
jupyter_client==7.4.9
jupyter_core==4.12.0
jupyter_server_fileid==0.8.0
jupyter_server_ydoc==0.6.1
jupyterlab==3.6.1
jupyterlab-pygments==0.2.2
jupyterlab_server==2.19.0
keras==2.11.0
libclang==15.0.6.1
Markdown==3.4.1
MarkupSafe==2.1.2
matplotlib-inline==0.1.6
mistune==2.0.5
nbclassic==0.5.2
nbclient==0.7.2
nbconvert==7.2.9
nbformat==5.7.3
nest-asyncio==1.5.6
notebook==6.5.2
notebook_shim==0.2.2
numpy==1.21.6
oauthlib==3.2.2
opencv-python==4.7.0.72
opt-einsum==3.3.0
packaging==23.0
pandas==1.3.5
pandocfilters==1.5.0
parso==0.8.3
pexpect==4.8.0
pickleshare==0.7.5
pkgutil_resolve_name==1.3.10
prometheus-client==0.16.0
prompt-toolkit==3.0.38
protobuf==3.19.6
psutil==5.9.4
ptyprocess==0.7.0
pyasn1==0.4.8
pyasn1-modules==0.2.8
pycparser==2.21
Pygments==2.14.0
pyhanlp==0.1.84
pyrsistent==0.19.3
PySocks==1.7.1
python-dateutil==2.8.2
python-json-logger==2.0.7
pytz==2022.7.1
PyYAML==6.0
pyzmq==25.0.0
regex==2022.10.31
requests==2.28.2
requests-oauthlib==1.3.1
rfc3339-validator==0.1.4
rfc3986-validator==0.1.1
rsa==4.9
Send2Trash==1.8.0
six==1.16.0
sklearn==0.0.post1
sniffio==1.3.0
soupsieve==2.4
tensorboard==2.11.2
tensorboard-data-server==0.6.1
tensorboard-plugin-wit==1.8.1
tensorflow==2.11.0
tensorflow-estimator==2.11.0
tensorflow-io-gcs-filesystem==0.31.0
termcolor==2.2.0
terminado==0.17.1
thulac==0.2.2
tinycss2==1.2.1
tokenizers==0.13.2
tomli==2.0.1
torch==1.13.1
tornado==6.2
tqdm==4.64.1
traitlets==5.9.0
transformers==4.26.1
typing_extensions==4.5.0
uri-template==1.2.0
urllib3==1.26.14
wcwidth==0.2.6
webcolors==1.12
webencodings==0.5.1
websocket-client==1.5.1
Werkzeug==2.2.3
wrapt==1.15.0
y-py==0.5.9
ypy-websocket==0.8.2
zipp==3.15.0

You May Also Enjoy