Update app.py

app.py

@@ -3,14 +3,14 @@ import torch.nn.functional as F
 from transformers import AutoTokenizer, AutoModel, AutoImageProcessor
 import gradio as gr
 import spaces
+import torch
 
 # neuralmind/bert-base-portuguese-cased
-ModelName = "papluca/xlm-roberta-base-language-detection"
+ModelName = "neuralmind/bert-base-portuguese-cased"
 model = AutoModel.from_pretrained(ModelName)
 tokenizer = AutoTokenizer.from_pretrained(ModelName, do_lower_case=False)
-
-# processor = AutoImageProcessor.from_pretrained("nomic-ai/nomic-embed-vision-v1.5")
-# vision_model = AutoModel.from_pretrained("nomic-ai/nomic-embed-vision-v1.5", trust_remote_code=True)
+processor = AutoImageProcessor.from_pretrained("nomic-ai/nomic-embed-vision-v1.5")
+vision_model = AutoModel.from_pretrained("nomic-ai/nomic-embed-vision-v1.5", trust_remote_code=True)
 
 # tokenizer = AutoTokenizer.from_pretrained('nomic-ai/nomic-embed-text-v1.5')
 # text_model = AutoModel.from_pretrained('nomic-ai/nomic-embed-text-v1.5', trust_remote_code=True)
@@ -23,7 +23,7 @@ def mean_pooling(model_output, attention_mask):
         
 @spaces.GPU
 def TxtEmbed(text):
-    import torch
+   
 
     
     input_ids = tokenizer.encode(text, return_tensors='pt')
@@ -31,20 +31,20 @@ def TxtEmbed(text):
     with torch.no_grad():
         outs = model(input_ids)
         encoded = outs[0][0, 1:-1]  # Ignore [CLS] and [SEP] special tokens
-
+    return (encoded.tolist())[0];
 
     
-   # sentences = [text]
-   # encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
+   #sentences = [text]
+   #encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
+   #
+   #with torch.no_grad():
+   #    model_output = text_model(**encoded_input)
+   #
+   #text_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
+   #text_embeddings = F.layer_norm(text_embeddings, normalized_shape=(text_embeddings.shape[1],))
+   #text_embeddings = F.normalize(text_embeddings, p=2, dim=1)
    # 
-   # with torch.no_grad():
-   #     model_output = text_model(**encoded_input)
-   # 
-   # text_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
-   # text_embeddings = F.layer_norm(text_embeddings, normalized_shape=(text_embeddings.shape[1],))
-   # text_embeddings = F.normalize(text_embeddings, p=2, dim=1)
-    
-    return (encoded.tolist())[0];
+   # return (text_embeddings.tolist)[0]