Twitter September 2021 (RoBERTa-base, 120M)
This is a RoBERTa-base model trained on 119.66M tweets until the end of September 2021. More details and performance scores are available in the TimeLMs paper.
Below, we provide some usage examples using the standard Transformers interface. For another interface more suited to comparing predictions and perplexity scores between models trained at different temporal intervals, check the TimeLMs repository.
For other models trained until different periods, check https://huggingface.co/cardiffnlp.
Preprocess Text
Replace usernames and links for placeholders: "@user" and "http". If you're interested in retaining verified users which were also retained during training, you may keep the users listed here.
def preprocess(text):
new_text = []
for t in text.split(" "):
t = '@user' if t.startswith('@') and len(t) > 1 else t
t = 'http' if t.startswith('http') else t
new_text.append(t)
return " ".join(new_text)
Example Masked Language Model
from transformers import pipeline, AutoTokenizer
MODEL = "cardiffnlp/twitter-roberta-base-sep2021"
fill_mask = pipeline("fill-mask", model=MODEL, tokenizer=MODEL)
tokenizer = AutoTokenizer.from_pretrained(MODEL)
def print_candidates():
for i in range(5):
token = tokenizer.decode(candidates[i]['token'])
score = candidates[i]['score']
print("%d) %.5f %s" % (i+1, score, token))
texts = [
"So glad I'm <mask> vaccinated.",
"I keep forgetting to bring a <mask>.",
"Looking forward to watching <mask> Game tonight!",
]
for text in texts:
t = preprocess(text)
print(f"{'-'*30}\n{t}")
candidates = fill_mask(t)
print_candidates()
Output:
------------------------------
So glad I'm <mask> vaccinated.
1) 0.39329 fully
2) 0.26694 getting
3) 0.17438 not
4) 0.03422 still
5) 0.01845 all
------------------------------
I keep forgetting to bring a <mask>.
1) 0.06773 mask
2) 0.04548 book
3) 0.03826 charger
4) 0.03506 backpack
5) 0.02997 bag
------------------------------
Looking forward to watching <mask> Game tonight!
1) 0.63009 the
2) 0.16154 The
3) 0.02110 this
4) 0.01903 End
5) 0.00810 Championship
Example Tweet Embeddings
from transformers import AutoTokenizer, AutoModel, TFAutoModel
import numpy as np
from scipy.spatial.distance import cosine
from collections import Counter
def get_embedding(text):
text = preprocess(text)
encoded_input = tokenizer(text, return_tensors='pt')
features = model(**encoded_input)
features = features[0].detach().cpu().numpy()
features_mean = np.mean(features[0], axis=0)
return features_mean
MODEL = "cardiffnlp/twitter-roberta-base-sep2021"
tokenizer = AutoTokenizer.from_pretrained(MODEL)
model = AutoModel.from_pretrained(MODEL)
query = "The book was awesome"
tweets = ["I just ordered fried chicken 🐣",
"The movie was great",
"What time is the next game?",
"Just finished reading 'Embeddings in NLP'"]
sims = Counter()
for tweet in tweets:
sim = 1 - cosine(get_embedding(query), get_embedding(tweet))
sims[tweet] = sim
print('Most similar to: ', query)
print(f"{'-'*30}")
for idx, (tweet, sim) in enumerate(sims.most_common()):
print("%d) %.5f %s" % (idx+1, sim, tweet))
Output:
Most similar to: The book was awesome
------------------------------
1) 0.99022 The movie was great
2) 0.96274 Just finished reading 'Embeddings in NLP'
3) 0.96006 I just ordered fried chicken 🐣
4) 0.95725 What time is the next game?
Example Feature Extraction
from transformers import AutoTokenizer, AutoModel, TFAutoModel
import numpy as np
MODEL = "cardiffnlp/twitter-roberta-base-sep2021"
tokenizer = AutoTokenizer.from_pretrained(MODEL)
text = "Good night 😊"
text = preprocess(text)
# Pytorch
model = AutoModel.from_pretrained(MODEL)
encoded_input = tokenizer(text, return_tensors='pt')
features = model(**encoded_input)
features = features[0].detach().cpu().numpy()
features_mean = np.mean(features[0], axis=0)
#features_max = np.max(features[0], axis=0)
# # Tensorflow
# model = TFAutoModel.from_pretrained(MODEL)
# encoded_input = tokenizer(text, return_tensors='tf')
# features = model(encoded_input)
# features = features[0].numpy()
# features_mean = np.mean(features[0], axis=0)
# #features_max = np.max(features[0], axis=0)