TimeLMs
Collection
Language Models specialised on social media, trained on different time periods.
โข
15 items
โข
Updated
This is a RoBERTa-base model trained on 123.86M tweets until the end of 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 this table.
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):
preprocessed_text = []
for t in text.split():
if len(t) > 1:
t = '@user' if t[0] == '@' and t.count('@') == 1 else t
t = 'http' if t.startswith('http') else t
preprocessed_text.append(t)
return ' '.join(preprocessed_text)
from transformers import pipeline, AutoTokenizer
MODEL = "cardiffnlp/twitter-roberta-base-2021-124m"
fill_mask = pipeline("fill-mask", model=MODEL, tokenizer=MODEL)
tokenizer = AutoTokenizer.from_pretrained(MODEL)
def pprint(candidates, n):
for i in range(n):
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)
pprint(candidates, 5)
Output:
------------------------------
So glad I'm <mask> vaccinated.
1) 0.39613 fully
2) 0.26333 getting
3) 0.18988 not
4) 0.02312 still
5) 0.02099 already
------------------------------
I keep forgetting to bring a <mask>.
1) 0.08356 mask
2) 0.05696 book
3) 0.03505 bag
4) 0.02983 backpack
5) 0.02847 blanket
------------------------------
Looking forward to watching <mask> Game tonight!
1) 0.46618 the
2) 0.24042 The
3) 0.03216 End
4) 0.02925 Squid
5) 0.02610 this
from transformers import AutoTokenizer, AutoModel, TFAutoModel
import numpy as np
from scipy.spatial.distance import cosine
from collections import Counter
def get_embedding(text): # naive approach for demonstration
text = preprocess(text)
encoded_input = tokenizer(text, return_tensors='pt')
features = model(**encoded_input)
features = features[0].detach().cpu().numpy()
return np.mean(features[0], axis=0)
MODEL = "cardiffnlp/twitter-roberta-base-2021-124m"
tokenizer = AutoTokenizer.from_pretrained(MODEL)
model = AutoModel.from_pretrained(MODEL)
query = "The book was awesome"
tweets = ["I just ordered fried chicken ๐ฃ",
"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) 0.98969 The movie was great
2) 0.96102 Just finished reading 'Embeddings in NLP'
3) 0.95565 I just ordered fried chicken ๐ฃ
4) 0.95041 What time is the next game?
from transformers import AutoTokenizer, AutoModel, TFAutoModel
import numpy as np
MODEL = "cardiffnlp/twitter-roberta-base-2021-124m"
tokenizer = AutoTokenizer.from_pretrained(MODEL)
text = "Good night ๐"
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)