Working on text to image

examples/text_to_image.py

@@ -2,6 +2,25 @@
 # This will be an image generator based on parquet data https://huggingface.co/datasets/yuvalkirstain/pickapic_v2/tree/main/data
 
 import pandas as pd
+from sentence_transformers import SentenceTransformer
+
+model = SentenceTransformer("all-MiniLM-L6-v2")
+
+# Our sentences we like to encode
+sentences = [
+    "This framework generates embeddings for each input sentence",
+    "Sentences are passed as a list of string.",
+    "The quick brown fox jumps over the lazy dog.",
+]
+
+# Sentences are encoded by calling model.encode()
+sentence_embeddings = model.encode(sentences)
+
+# Print the embeddings
+for sentence, embedding in zip(sentences, sentence_embeddings):
+    print("Sentence:", sentence)
+    print("Embedding:", len(embedding))
+    print("")
 
 ans = pd.read_parquet("train-00000-of-00645-b66ac786bf6fb553.parquet")
 print('ans', ans)

pyproject.toml

@@ -20,6 +20,7 @@ torchvision = "^0.16.2"
 pytorch-lightning = "^2.1.3"
 pandas = "^2.2.2"
 fastparquet = "^2024.2.0"
+sentence-transformers = "^2.7.0"
 
 [tool.poetry.dev-dependencies]
 black = "^22.3.0"