Edit model card


This is a XLM-Roberta-base model trained on ~198M multilingual tweets, described and evaluated in the reference paper. To evaluate this and other LMs on Twitter-specific data, please refer to the main repository. A usage example is provided below.

Computing tweet similarity

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
    return " ".join(new_text)

def get_embedding(text):
    text = preprocess(text)
    encoded_input = tokenizer(text, return_tensors='pt')
    features = model(**encoded_input)
    features = features[0].detach().numpy() 
    features_mean = np.mean(features[0], axis=0) 
    return features_mean

query = "Acabo de pedir pollo frito 🐣" #spanish

tweets = ["We had a great time! ⚽️", # english
          "We hebben een geweldige tijd gehad! β›©", # dutch
          "Nous avons passΓ© un bon moment! πŸŽ₯", # french
          "Ci siamo divertiti! 🍝"] # italian

d = defaultdict(int)
for tweet in tweets:
    sim = 1-cosine(get_embedding(query),get_embedding(tweet))
    d[tweet] = sim
print('Most similar to: ',query)
for idx,x in enumerate(sorted(d.items(), key=lambda x:x[1], reverse=True)):
Most similar to:  Acabo de pedir pollo frito 🐣
1 Ci siamo divertiti! 🍝
2 Nous avons passΓ© un bon moment! πŸŽ₯
3 We had a great time! ⚽️
4 We hebben een geweldige tijd gehad! β›©
Downloads last month
Hosted inference API
Mask token: <mask>
This model can be loaded on the Inference API on-demand.