changed num workers

app.py

@@ -16,14 +16,24 @@ def process_file(file, model_name):
     shutil.copyfile(file.name, saved_test_dataset)
     # For demonstration purposes, we'll just return the content with the selected model name
     subprocess.run(["python", "src/test_saved_model.py"])
-    return f"Model: {model_name}\nContent:\n{content}"
+    result = {}
+    with open("result.txt", 'r') as file:
+        for line in file:
+            key, value = line.strip().split(': ', 1)
+            # print(type(key))
+            if key=='epoch':
+                result[key]=value
+            else:
+                 result[key]=float(value)
+
+    return f"Model: {model_name}\nResult:\n{result}"
 
 # List of models for the dropdown menu
 models = ["Model A", "Model B", "Model C"]
 
 # Create the Gradio interface
 with gr.Blocks() as demo:
-    gr.Markdown("# File Processor with Model Selection")
+    gr.Markdown("# ASTRA")
     gr.Markdown("Upload a .txt file and select a model from the dropdown menu.")
     
     with gr.Row():

result.txt

@@ -0,0 +1,6 @@
+epoch: EP0_test
+accuracy: 12.545819442371167
+avg_loss: 0.0
+precisions: 0.9988672445640735
+recalls: 0.8782073609659816
+f1_scores: 0.9254850123297983

src/test_saved_model.py

@@ -148,7 +148,9 @@ class BERTFineTunedTrainer:
             "recalls": recalls,
             "f1_scores": f1_scores
         }
-        
+        with open("result.txt", 'w') as file:
+            for key, value in final_msg.items():
+                file.write(f"{key}: {value}\n")
         print(final_msg)
 
             # print("EP%d_%s, avg_loss=" % (epoch, str_code), avg_loss / len(data_iter), "total_acc=", total_correct * 100.0 / total_element)
@@ -217,7 +219,7 @@ if __name__ == "__main__":
     print("Loading Test Dataset", args.test_dataset)
     test_dataset = TokenizerDataset(args.test_dataset, args.test_label, vocab_obj, seq_len=args.seq_len, train=False)
     print("Creating Dataloader")
-    test_data_loader = DataLoader(test_dataset, batch_size=args.batch_size, num_workers=0)
+    test_data_loader = DataLoader(test_dataset, batch_size=args.batch_size, num_workers=4)
     bert = torch.load(args.finetuned_bert_checkpoint, map_location="cpu")
         
     if args.workspace_name == "ratio_proportion_change4":