writing output to logs

app.py

@@ -5,6 +5,7 @@ import requests
 import matplotlib.pyplot as plt
 import gradio as gr
 from transformers import ImageClassificationPipeline, PerceiverForImageClassificationConvProcessing, PerceiverFeatureExtractor
+from transformers import AutoTokenizer
 import torch
 
 
@@ -25,8 +26,27 @@ model = PerceiverForImageClassificationConvProcessing.from_pretrained("deepmind/
 
 image_pipe = ImageClassificationPipeline(model=model, feature_extractor=feature_extractor)
 
+'''
+repo_name = "ydshieh/vit-gpt2-coco-en"
+
+feature_extractor2 = ViTFeatureExtractor.from_pretrained(repo_name)
+tokenizer = AutoTokenizer.from_pretrained(repo_name)
+model2 = VisionEncoderDecoderModel.from_pretrained(repo_name)
+pixel_values = feature_extractor2(image, return_tensors="pt").pixel_values
+
+# autoregressively generate text (using beam search or other decoding strategy)
+generated_ids = model2.generate(pixel_values, max_length=16, num_beams=4, return_dict_in_generate=True)
+# decode into text
+preds = tokenizer.batch_decode(generated_ids[0], skip_special_tokens=True)
+preds = [pred.strip() for pred in preds]
+print(preds)
+'''
+
+
 def classify_image(image):
   results = image_pipe(image)
+  
+  print("RESULTS")
   print(results)
   # convert to format Gradio expects
   output = {}
@@ -34,6 +54,8 @@ def classify_image(image):
     predicted_label = prediction['label']
     score = prediction['score']
     output[predicted_label] = score
+  print("OUTPUT")
+  print(output)
   return output