 View on TensorFlow.org
|
 Run in Google Colab
|
 View on GitHub
|
 Download notebook
|
 See TF Hub model
|
The CORD-19 Swivel text embedding module from TF-Hub (https://tfhub.dev/tensorflow/cord-19/swivel-128d/1) was built to support researchers analyzing natural languages text related to COVID-19. These embeddings were trained on the titles, authors, abstracts, body texts, and reference titles of articles in the CORD-19 dataset.
In this colab we will:
- Analyze semantically similar words in the embedding space
- Train a classifier on the SciCite dataset using the CORD-19 embeddings
Setup
import functools
import itertools
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
import pandas as pd
import tensorflow.compat.v1 as tf
tf.disable_eager_execution()
tf.logging.set_verbosity('ERROR')
import tensorflow_datasets as tfds
import tensorflow_hub as hub
try:
from google.colab import data_table
def display_df(df):
return data_table.DataTable(df, include_index=False)
except ModuleNotFoundError:
# If google-colab is not available, just display the raw DataFrame
def display_df(df):
return df
Analyze the embeddings
Let's start off by analyzing the embedding by calculating and plotting a correlation matrix between different terms. If the embedding learned to successfully capture the meaning of different words, the embedding vectors of semantically similar words should be close together. Let's take a look at some COVID-19 related terms.
# Use the inner product between two embedding vectors as the similarity measure
def plot_correlation(labels, features):
corr = np.inner(features, features)
corr /= np.max(corr)
sns.heatmap(corr, xticklabels=labels, yticklabels=labels)
with tf.Graph().as_default():
# Load the module
query_input = tf.placeholder(tf.string)
module = hub.Module('https://tfhub.dev/tensorflow/cord-19/swivel-128d/1')
embeddings = module(query_input)
with tf.train.MonitoredTrainingSession() as sess:
# Generate embeddings for some terms
queries = [
# Related viruses
"coronavirus", "SARS", "MERS",
# Regions
"Italy", "Spain", "Europe",
# Symptoms
"cough", "fever", "throat"
]
features = sess.run(embeddings, feed_dict={query_input: queries})
plot_correlation(queries, features)
2022-06-23 11:07:42.304283: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions.
We can see that the embedding successfully captured the meaning of the different terms. Each word is similar to the other words of its cluster (i.e. "coronavirus" highly correlates with "SARS" and "MERS"), while they are different from terms of other clusters (i.e. the similarity between "SARS" and "Spain" is close to 0).
Now let's see how we can use these embeddings to solve a specific task.
SciCite: Citation Intent Classification
This section shows how one can use the embedding for downstream tasks such as text classification. We'll use the SciCite dataset from TensorFlow Datasets to classify citation intents in academic papers. Given a sentence with a citation from an academic paper, classify whether the main intent of the citation is as background information, use of methods, or comparing results.
Set up the dataset from TFDS
class Dataset:
"""Build a dataset from a TFDS dataset."""
def __init__(self, tfds_name, feature_name, label_name):
self.dataset_builder = tfds.builder(tfds_name)
self.dataset_builder.download_and_prepare()
self.feature_name = feature_name
self.label_name = label_name
def get_data(self, for_eval):
splits = THE_DATASET.dataset_builder.info.splits
if tfds.Split.TEST in splits:
split = tfds.Split.TEST if for_eval else tfds.Split.TRAIN
else:
SPLIT_PERCENT = 80
split = "train[{}%:]".format(SPLIT_PERCENT) if for_eval else "train[:{}%]".format(SPLIT_PERCENT)
return self.dataset_builder.as_dataset(split=split)
def num_classes(self):
return self.dataset_builder.info.features[self.label_name].num_classes
def class_names(self):
return self.dataset_builder.info.features[self.label_name].names
def preprocess_fn(self, data):
return data[self.feature_name], data[self.label_name]
def example_fn(self, data):
feature, label = self.preprocess_fn(data)
return {'feature': feature, 'label': label}, label
def get_example_data(dataset, num_examples, **data_kw):
"""Show example data"""
with tf.Session() as sess:
batched_ds = dataset.get_data(**data_kw).take(num_examples).map(dataset.preprocess_fn).batch(num_examples)
it = tf.data.make_one_shot_iterator(batched_ds).get_next()
data = sess.run(it)
return data
TFDS_NAME = 'scicite'
TEXT_FEATURE_NAME = 'string'
LABEL_NAME = 'label'
THE_DATASET = Dataset(TFDS_NAME, TEXT_FEATURE_NAME, LABEL_NAME)
Let's take a look at a few labeled examples from the training set
NUM_EXAMPLES = 20
data = get_example_data(THE_DATASET, NUM_EXAMPLES, for_eval=False)
display_df(
pd.DataFrame({
TEXT_FEATURE_NAME: [ex.decode('utf8') for ex in data[0]],
LABEL_NAME: [THE_DATASET.class_names()[x] for x in data[1]]
}))
Training a citaton intent classifier
We'll train a classifier on the SciCite dataset using an Estimator. Let's set up the input_fns to read the dataset into the model
def preprocessed_input_fn(for_eval):
data = THE_DATASET.get_data(for_eval=for_eval)
data = data.map(THE_DATASET.example_fn, num_parallel_calls=1)
return data
def input_fn_train(params):
data = preprocessed_input_fn(for_eval=False)
data = data.repeat(None)
data = data.shuffle(1024)
data = data.batch(batch_size=params['batch_size'])
return data
def input_fn_eval(params):
data = preprocessed_input_fn(for_eval=True)
data = data.repeat(1)
data = data.batch(batch_size=params['batch_size'])
return data
def input_fn_predict(params):
data = preprocessed_input_fn(for_eval=True)
data = data.batch(batch_size=params['batch_size'])
return data
Let's build a model which use the CORD-19 embeddings with a classification layer on top.
def model_fn(features, labels, mode, params):
# Embed the text
embed = hub.Module(params['module_name'], trainable=params['trainable_module'])
embeddings = embed(features['feature'])
# Add a linear layer on top
logits = tf.layers.dense(
embeddings, units=THE_DATASET.num_classes(), activation=None)
predictions = tf.argmax(input=logits, axis=1)
if mode == tf.estimator.ModeKeys.PREDICT:
return tf.estimator.EstimatorSpec(
mode=mode,
predictions={
'logits': logits,
'predictions': predictions,
'features': features['feature'],
'labels': features['label']
})
# Set up a multi-class classification head
loss = tf.nn.sparse_softmax_cross_entropy_with_logits(
labels=labels, logits=logits)
loss = tf.reduce_mean(loss)
if mode == tf.estimator.ModeKeys.TRAIN:
optimizer = tf.train.GradientDescentOptimizer(learning_rate=params['learning_rate'])
train_op = optimizer.minimize(loss, global_step=tf.train.get_or_create_global_step())
return tf.estimator.EstimatorSpec(mode=mode, loss=loss, train_op=train_op)
elif mode == tf.estimator.ModeKeys.EVAL:
accuracy = tf.metrics.accuracy(labels=labels, predictions=predictions)
precision = tf.metrics.precision(labels=labels, predictions=predictions)
recall = tf.metrics.recall(labels=labels, predictions=predictions)
return tf.estimator.EstimatorSpec(
mode=mode,
loss=loss,
eval_metric_ops={
'accuracy': accuracy,
'precision': precision,
'recall': recall,
})
Hyperparmeters
EMBEDDING = 'https://tfhub.dev/tensorflow/cord-19/swivel-128d/1'
TRAINABLE_MODULE = False
STEPS = 8000
EVAL_EVERY = 200
BATCH_SIZE = 10
LEARNING_RATE = 0.01
params = {
'batch_size': BATCH_SIZE,
'learning_rate': LEARNING_RATE,
'module_name': EMBEDDING,
'trainable_module': TRAINABLE_MODULE
}
Train and evaluate the model
Let's train and evaluate the model to see the performance on the SciCite task
estimator = tf.estimator.Estimator(functools.partial(model_fn, params=params))
metrics = []
for step in range(0, STEPS, EVAL_EVERY):
estimator.train(input_fn=functools.partial(input_fn_train, params=params), steps=EVAL_EVERY)
step_metrics = estimator.evaluate(input_fn=functools.partial(input_fn_eval, params=params))
print('Global step {}: loss {:.3f}, accuracy {:.3f}'.format(step, step_metrics['loss'], step_metrics['accuracy']))
metrics.append(step_metrics)
2022-06-23 11:07:46.960470: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. /tmpfs/tmp/ipykernel_7976/393120678.py:7: UserWarning: `tf.layers.dense` is deprecated and will be removed in a future version. Please use `tf.keras.layers.Dense` instead. logits = tf.layers.dense( 2022-06-23 11:07:49.557452: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 0: loss 0.800, accuracy 0.676 2022-06-23 11:07:50.725489: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:07:52.061311: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 200: loss 0.717, accuracy 0.710 2022-06-23 11:07:52.925683: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:07:54.158943: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 400: loss 0.673, accuracy 0.737 2022-06-23 11:07:55.013043: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:07:56.222225: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 600: loss 0.640, accuracy 0.750 2022-06-23 11:07:57.134483: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:07:58.332311: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 800: loss 0.621, accuracy 0.761 2022-06-23 11:07:59.199464: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:00.417410: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 1000: loss 0.616, accuracy 0.756 2022-06-23 11:08:01.283511: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:02.844535: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 1200: loss 0.602, accuracy 0.769 2022-06-23 11:08:03.740984: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:04.963629: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 1400: loss 0.585, accuracy 0.778 2022-06-23 11:08:05.834220: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:07.053425: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 1600: loss 0.593, accuracy 0.767 2022-06-23 11:08:07.926544: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:09.159102: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 1800: loss 0.581, accuracy 0.774 2022-06-23 11:08:10.010124: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:11.240685: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 2000: loss 0.564, accuracy 0.786 2022-06-23 11:08:12.121468: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:13.369059: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 2200: loss 0.564, accuracy 0.783 2022-06-23 11:08:14.297963: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:15.591648: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 2400: loss 0.563, accuracy 0.781 2022-06-23 11:08:16.461064: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:17.677507: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 2600: loss 0.560, accuracy 0.787 2022-06-23 11:08:18.575776: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:19.849815: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 2800: loss 0.555, accuracy 0.796 2022-06-23 11:08:20.738021: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:22.028718: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 3000: loss 0.560, accuracy 0.783 2022-06-23 11:08:22.919569: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:24.168511: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 3200: loss 0.550, accuracy 0.788 2022-06-23 11:08:25.058172: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:26.545155: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 3400: loss 0.553, accuracy 0.790 2022-06-23 11:08:27.458234: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:28.695293: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 3600: loss 0.554, accuracy 0.782 2022-06-23 11:08:29.564570: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:30.822945: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 3800: loss 0.557, accuracy 0.776 2022-06-23 11:08:31.694732: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:32.931036: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 4000: loss 0.546, accuracy 0.795 2022-06-23 11:08:33.841319: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:35.053802: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 4200: loss 0.548, accuracy 0.788 2022-06-23 11:08:35.902580: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:37.117769: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 4400: loss 0.547, accuracy 0.793 2022-06-23 11:08:37.983478: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:39.246484: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 4600: loss 0.546, accuracy 0.791 2022-06-23 11:08:40.096748: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:41.304329: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 4800: loss 0.549, accuracy 0.779 2022-06-23 11:08:42.152740: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:43.363747: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 5000: loss 0.543, accuracy 0.795 2022-06-23 11:08:44.208074: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:45.459051: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 5200: loss 0.538, accuracy 0.799 2022-06-23 11:08:46.327111: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:47.554322: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 5400: loss 0.553, accuracy 0.779 2022-06-23 11:08:48.403682: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:49.696784: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 5600: loss 0.541, accuracy 0.798 2022-06-23 11:08:50.802686: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:52.049043: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 5800: loss 0.542, accuracy 0.795 2022-06-23 11:08:52.915377: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:54.122929: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 6000: loss 0.537, accuracy 0.797 2022-06-23 11:08:54.977123: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:56.166756: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 6200: loss 0.545, accuracy 0.783 2022-06-23 11:08:57.016505: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:08:58.253463: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 6400: loss 0.550, accuracy 0.785 2022-06-23 11:08:59.126682: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:09:00.318613: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 6600: loss 0.535, accuracy 0.794 2022-06-23 11:09:01.185146: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:09:02.391791: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 6800: loss 0.541, accuracy 0.793 2022-06-23 11:09:03.241711: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:09:04.496313: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 7000: loss 0.539, accuracy 0.795 2022-06-23 11:09:05.361386: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:09:06.590139: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 7200: loss 0.550, accuracy 0.781 2022-06-23 11:09:07.491842: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:09:08.711150: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 7400: loss 0.536, accuracy 0.791 2022-06-23 11:09:09.580727: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:09:10.813533: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 7600: loss 0.533, accuracy 0.799 2022-06-23 11:09:11.661249: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. 2022-06-23 11:09:12.873331: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. Global step 7800: loss 0.539, accuracy 0.789
global_steps = [x['global_step'] for x in metrics]
fig, axes = plt.subplots(ncols=2, figsize=(20,8))
for axes_index, metric_names in enumerate([['accuracy', 'precision', 'recall'],
['loss']]):
for metric_name in metric_names:
axes[axes_index].plot(global_steps, [x[metric_name] for x in metrics], label=metric_name)
axes[axes_index].legend()
axes[axes_index].set_xlabel("Global Step")
We can see that the loss quickly decreases while especially the accuracy rapidly increases. Let's plot some examples to check how the prediction relates to the true labels:
predictions = estimator.predict(functools.partial(input_fn_predict, params))
first_10_predictions = list(itertools.islice(predictions, 10))
display_df(
pd.DataFrame({
TEXT_FEATURE_NAME: [pred['features'].decode('utf8') for pred in first_10_predictions],
LABEL_NAME: [THE_DATASET.class_names()[pred['labels']] for pred in first_10_predictions],
'prediction': [THE_DATASET.class_names()[pred['predictions']] for pred in first_10_predictions]
}))
2022-06-23 11:09:14.223233: W tensorflow/core/common_runtime/graph_constructor.cc:1526] Importing a graph with a lower producer version 27 into an existing graph with producer version 1087. Shape inference will have run different parts of the graph with different producer versions. /tmpfs/tmp/ipykernel_7976/393120678.py:7: UserWarning: `tf.layers.dense` is deprecated and will be removed in a future version. Please use `tf.keras.layers.Dense` instead. logits = tf.layers.dense(
We can see that for this random sample, the model predicts the correct label most of the times, indicating that it can embed scientific sentences pretty well.
What's next?
Now that you've gotten to know a bit more about the CORD-19 Swivel embeddings from TF-Hub, we encourage you to participate in the CORD-19 Kaggle competition to contribute to gaining scientific insights from COVID-19 related academic texts.
- Participate in the CORD-19 Kaggle Challenge
- Learn more about the COVID-19 Open Research Dataset (CORD-19)
- See documentation and more about the TF-Hub embeddings at https://tfhub.dev/tensorflow/cord-19/swivel-128d/1
- Explore the CORD-19 embedding space with the TensorFlow Embedding Projector