Loading SBERT_Model.py +26 −0 Changes for SBERT_Model.py: 26 added lines, 0 removed lines. Original line number Diff line number Diff line Loading @@ -4,6 +4,11 @@ import STSB_Dataset from scipy.stats import spearmanr, pearsonr from sklearn import metrics import numpy as np import scipy.spatial as sp from sklearn.decomposition import PCA import matplotlib.pyplot as plt from random import randrange class SBERT_Model: def __init__(self, name, filepath, dataset, probing=False): Loading @@ -20,6 +25,27 @@ class SBERT_Model: self.cosine_scores = self.get_cosine_scores() self.preds = self.get_preds() def visualize_embeddings(self): sentences = self.sentences1 + self.sentences2 embs = self.get_embeddings(sentences) pca = PCA(n_components=2) X = pca.fit_transform(embs_together) plt.figure(figsize=(20,10)) plt.scatter(X[:, 0], X[:, 1]) for x, y in X: label = "{:.2f}".format(y) name = i label = f"{name}\n({x:.2f},{y:.2f})" plt.annotate(label, # this is the text (x,y), # this is the point to label textcoords="offset points", # how to position the text xytext=(0,10), # distance from text to points (x,y) ha='center') # horizontal alignment can be left, right or center i = i + 1 plt.savefig(f'{randrange(0,10000)}plot.png') def get_embeddings(self, sentences): return self.model.encode(sentences, convert_to_tensor=True) Loading Loading
SBERT_Model.py +26 −0 Changes for SBERT_Model.py: 26 added lines, 0 removed lines. Original line number Diff line number Diff line Loading @@ -4,6 +4,11 @@ import STSB_Dataset from scipy.stats import spearmanr, pearsonr from sklearn import metrics import numpy as np import scipy.spatial as sp from sklearn.decomposition import PCA import matplotlib.pyplot as plt from random import randrange class SBERT_Model: def __init__(self, name, filepath, dataset, probing=False): Loading @@ -20,6 +25,27 @@ class SBERT_Model: self.cosine_scores = self.get_cosine_scores() self.preds = self.get_preds() def visualize_embeddings(self): sentences = self.sentences1 + self.sentences2 embs = self.get_embeddings(sentences) pca = PCA(n_components=2) X = pca.fit_transform(embs_together) plt.figure(figsize=(20,10)) plt.scatter(X[:, 0], X[:, 1]) for x, y in X: label = "{:.2f}".format(y) name = i label = f"{name}\n({x:.2f},{y:.2f})" plt.annotate(label, # this is the text (x,y), # this is the point to label textcoords="offset points", # how to position the text xytext=(0,10), # distance from text to points (x,y) ha='center') # horizontal alignment can be left, right or center i = i + 1 plt.savefig(f'{randrange(0,10000)}plot.png') def get_embeddings(self, sentences): return self.model.encode(sentences, convert_to_tensor=True) Loading