tweet-snest / app.py
wilmerags's picture
feat: Adapting perplexity given smaller datasets are expected
dd9138c
raw
history blame
4.65 kB
from typing import List
import numpy as np
import streamlit as st
import tweepy
import hdbscan
from bokeh.models import ColumnDataSource, HoverTool
from bokeh.palettes import Cividis256 as Pallete
from bokeh.plotting import Figure, figure
from bokeh.transform import factor_cmap
from sklearn.manifold import TSNE
from sentence_transformers import SentenceTransformer
client = tweepy.Client(bearer_token=st.secrets["tw_bearer_token"])
model_to_use = {
"English": "all-MiniLM-L12-v2",
"Use all the ones you know (~15 lang)": "paraphrase-multilingual-MiniLM-L12-v2"
}
# Original implementation from: https://huggingface.co/spaces/edugp/embedding-lenses/blob/main/app.py
SEED = 42
@st.cache(show_spinner=False, allow_output_mutation=True)
def load_model(model_name: str) -> SentenceTransformer:
embedder = model_name
return SentenceTransformer(embedder)
def embed_text(text: List[str], model: SentenceTransformer) -> np.ndarray:
return model.encode(text)
def get_tsne_embeddings(
embeddings: np.ndarray, perplexity: int = 15, n_components: int = 2, init: str = "pca", n_iter: int = 5000, random_state: int = SEED
) -> np.ndarray:
tsne = TSNE(perplexity=perplexity, n_components=n_components, init=init, n_iter=n_iter, random_state=random_state)
return tsne.fit_transform(embeddings)
def draw_interactive_scatter_plot(
texts: np.ndarray, xs: np.ndarray, ys: np.ndarray, values: np.ndarray, labels: np.ndarray, text_column: str, label_column: str
) -> Figure:
# Normalize values to range between 0-255, to assign a color for each value
max_value = values.max()
min_value = values.min()
if max_value - min_value == 0:
values_color = np.ones(len(values))
else:
values_color = ((values - min_value) / (max_value - min_value) * 255).round().astype(int).astype(str)
values_color_set = sorted(values_color)
values_list = values.astype(str).tolist()
values_set = sorted(values_list)
labels_list = labels.astype(str).tolist()
source = ColumnDataSource(data=dict(x=xs, y=ys, text=texts, label=values_list, original_label=labels_list))
hover = HoverTool(tooltips=[(text_column, "@text{safe}"), (label_column, "@original_label")])
p = figure(plot_width=800, plot_height=800, tools=[hover])
p.circle("x", "y", size=10, source=source, fill_color=factor_cmap("label", palette=[Pallete[int(id_)] for id_ in values_color_set], factors=values_set))
p.axis.visible = False
p.xgrid.grid_line_color = None
p.ygrid.grid_line_color = None
p.toolbar.logo = None
return p
# Up to here
def generate_plot(
df: List[str],
model: SentenceTransformer,
) -> Figure:
with st.spinner(text="Embedding text..."):
embeddings = embed_text(df, model)
# encoded_labels = encode_labels(labels)
cluster = hdbscan.HDBSCAN(
min_cluster_size=5,
metric='euclidean',
cluster_selection_method='eom'
).fit(embeddings)
encoded_labels = cluster.labels_
with st.spinner("Reducing dimensionality..."):
embeddings_2d = get_tsne_embeddings(embeddings)
plot = draw_interactive_scatter_plot(
df, embeddings_2d[:, 0], embeddings_2d[:, 1], encoded_labels, encoded_labels, 'text', 'label'
)
return plot
st.title("Tweet-SNEst")
st.write("Visualize tweets embeddings in 2D using colors for topics labels.")
col1, col2 = st.columns(2)
with col1:
tw_user = st.text_input("Twitter handle", "huggingface")
with col2:
tw_sample = st.number_input("Maximum number of tweets to use", 1, 300, 100, 10)
expected_lang = st.radio(
"What language should be assumed to be found?",
('English', 'Use all the ones you know (~15 lang)'),
0
)
with st.spinner(text="Loading model..."):
model = load_model(model_to_use[expected_lang])
usr = client.get_user(username=tw_user)
# st.write(usr.data.id)
if tw_user:
with st.spinner(f"Getting to know the '{tw_user}'..."):
tweets_objs = []
while tw_sample >= 100:
current_sample = min(100, tw_sample)
tweets_response = client.get_users_tweets(usr.data.id, max_results=current_sample)
tweets_objs += tweets_response.data
tw_sample -= current_sample
if tw_sample > 0:
tweets_response = client.get_users_tweets(usr.data.id, max_results=tw_sample)
tweets_objs += tweets_response.data
tweets_txt = [tweet.text for tweet in tweets_objs]
# plot = generate_plot(df, text_column, label_column, sample, dimensionality_reduction_function, model)
plot = generate_plot(tweets_txt, model)
st.bokeh_chart(plot)