Update app.py

app.py

@@ -14,7 +14,7 @@ df = pd.read_csv(url)
 
 # Tokenizer
 class ScratchTokenizer:
-    def _init_(self):
+    def __init__(self):
         self.word2idx = {"<PAD>": 0, "<SOS>": 1, "<EOS>": 2, "<UNK>": 3}
         self.idx2word = {0: "<PAD>", 1: "<SOS>", 2: "<EOS>", 3: "<UNK>"}
         self.vocab_size = 4
@@ -44,15 +44,15 @@ tokenizer.build_vocab(train_data["instruction"].tolist() + train_data["response"
 
 # Dataset Class (not used in inference but useful for training)
 class TextDataset(Dataset):
-    def _init_(self, data, tokenizer, max_len=200):
+    def __init__(self, data, tokenizer, max_len=200):
         self.data = data
         self.tokenizer = tokenizer
         self.max_len = max_len
 
-    def _len_(self):
+    def __len__(self):
         return len(self.data)
 
-    def _getitem_(self, idx):
+    def __getitem__(self, idx):
         src_text = self.data.iloc[idx]["instruction"]
         tgt_text = self.data.iloc[idx]["response"]
         src = torch.tensor(self.tokenizer.encode(src_text), dtype=torch.long)
@@ -61,8 +61,8 @@ class TextDataset(Dataset):
 
 # Model
 class GPTModel(nn.Module):
-    def _init_(self, vocab_size, embed_size=256, num_heads=8, num_layers=6, max_len=200):
-        super(GPTModel, self)._init_()
+    def __init__(self, vocab_size, embed_size=256, num_heads=8, num_layers=6, max_len=200):
+        super(GPTModel, self).__init__()
         self.embedding = nn.Embedding(vocab_size, embed_size)
         self.pos_embedding = nn.Parameter(torch.randn(1, max_len, embed_size))
         self.transformer = nn.TransformerDecoder(