Update app.py

app.py

@@ -1,7 +1,187 @@
 import gradio as gr
 
-def greet(name):
-    return "Hello " + name + "!!"
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
+batch_size = 64 # how many independent sequences will we process in parallel?
+block_size = 256 # what is the maximum context length for predictions?
+max_iters = 5000
+eval_interval = 500
+learning_rate = 3e-4
+device = 'cuda' if torch.cuda.is_available() else 'cpu'
+print(f"The code is running on {device} : GPU={torch.cuda.get_device_name(0)}")
+eval_iters = 200
+n_embd = 384
+n_head = 6
+n_layer = 6
+dropout = 0.2
+
+
+torch.manual_seed(1337)
+
+# wget https://raw.githubusercontent.com/karpathy/char-rnn/master/data/tinyshakespeare/input.txt
+with open('input.txt', 'r', encoding='utf-8') as f:
+    text = f.read()
+
+# here are all the unique characters that occur in this text
+chars = sorted(list(set(text)))
+vocab_size = len(chars)
+# create a mapping from characters to integers
+stoi = { ch:i for i,ch in enumerate(chars) }
+itos = { i:ch for i,ch in enumerate(chars) }
+encode = lambda s: [stoi[c] for c in s] # encoder: take a string, output a list of integers
+decode = lambda l: ''.join([itos[i] for i in l]) # decoder: take a list of integers, output a string
+
+        
+class Head(nn.Module):
+    """ one head of self-attention """
+    
+    def __init__(self, head_size):
+        super().__init__()
+        self.key = nn.Linear(n_embd, head_size, bias=False)
+        self.query = nn.Linear(n_embd, head_size, bias=False)
+        self.value = nn.Linear(n_embd, head_size, bias=False)
+        self.register_buffer('tril', torch.tril(torch.ones(block_size, block_size))) # create lower triangular matrix
+        
+        self.dropout = nn.Dropout(dropout)
+    
+    def forward(self, x):
+        B,T,C = x.shape
+        k = self.key(x) # B, T, C
+        q = self.query(x) # B, T, C
+        # compute attention scores = ("affinities")
+        wei = q @ k.transpose(-2, -1) * C**-0.5 # (B, T, C) @ (B, C, T) -> (B, T, T)
+        #wei = wei.masked_fill(self.tril[:T, :T]==0, float('-inf')) # (B, T, T)
+        tril = torch.tril(torch.ones(T, T)).to(device)
+        wei = wei.masked_fill(tril == 0, float('-inf'))
+        wei = F.softmax(wei, dim=-1) # (B, T, T)
+        wei = self.dropout(wei)
+        # perform the weighted aggregation of the values
+        v = self.value(x) # (B, T, C)
+        out = wei @ v
+        return out
+        
+        
+class MultiHeadAttention(nn.Module):
+    """ multiple heads of self-attention in parallel """
+    
+    def __init__(self, num_heads, head_size):
+        super().__init__()
+        self.heads = nn.ModuleList([Head(head_size) for _ in range(num_heads)])
+        self.proj = nn.Linear(n_embd, n_embd)
+        self.dropout = nn.Dropout(dropout)
+    
+    def forward(self, x):
+        out = torch.cat([h(x) for h in self.heads], dim=-1) # h(x) call forward function is Head class
+        out = self.dropout(self.proj(out))
+        return out
+
+class FeedForward(nn.Module): # per token level, every token does this independently, its allowing tokens to think on data provided by self attention
+    """ a simple linear layer followed by a non-linearity"""
+    
+    def __init__(self, n_embd):
+        super().__init__()
+        self.net = nn.Sequential(
+            nn.Linear(n_embd, 4 * n_embd), # we multiply by 4 cause the paper says so 
+            nn.ReLU(),
+            nn.Linear(4 * n_embd, n_embd),
+            nn.Dropout(dropout)
+        )
+    
+    def forward(self, x):
+        return self.net(x)
+
+class Block(nn.Module):
+    """Transformer block: communication followed by computation """
+    
+    def __init__(self, n_embed, n_head):
+        # n_embd: embedding dimension, n_head: the number of heads we'd like
+        super().__init__()
+        head_size = n_embd // n_head
+        self.sa = MultiHeadAttention(n_head, head_size)
+        self.ffwd = FeedForward(n_embd)
+        self.ln1 = nn.LayerNorm(n_embd)
+        self.ln2 = nn.LayerNorm(n_embd)
+        
+    def forward(self, x):
+        x = x + self.sa(self.ln1(x)) # x = x + self .. is residual connection
+        x = x + self.ffwd(self.ln2(x))
+        return x
+        
+
+class BigramLanguageModel(nn.Module):
+
+    def __init__(self):
+        super().__init__()
+        # each token directly reads off the logits for the next token from a lookup table
+        self.token_embedding_table = nn.Embedding(vocab_size, n_embd)
+        self.position_embedding_table = nn.Embedding(block_size, n_embd) # so each position from 0 to block_size - 1 will also get its own embedding vector
+        self.blocks = nn.Sequential(*[Block(n_embd, n_head=n_head) for _ in range(n_layer)])
+        self.ln_f = nn.LayerNorm(n_embd) # final layer Norm
+        self.lm_head = nn.Linear(n_embd, vocab_size)
+
+    def forward(self, idx, targets=None):
+        B, T = idx.shape
+        
+        # idx and targets are both (B,T) tensor of integers
+        tok_emb = self.token_embedding_table(idx) # (B,T,C=n_embed)
+        pos_emb = self.position_embedding_table(torch.arange(T, device=device)) # (T, C)
+        # pos_emb tensor will be a (block_size, n_emb) tensor # block_size is max context length for predictions
+        # each row represents the embedding vector for the corresponding position
+        # so 0th row will represent the vector for 0th position
+        x = tok_emb + pos_emb # (B, T, C)
+        x = self.blocks(x) # (B, T, C)
+        logits = self.lm_head(x) # (B, T, C=vocab_size)
+
+        if targets is None:
+            loss = None
+        else:
+            B, T, C = logits.shape
+            logits = logits.view(B*T, C)
+            targets = targets.view(B*T)
+            loss = F.cross_entropy(logits, targets)
+
+        return logits, loss
+
+    def generate(self, idx, max_new_tokens):
+        # idx is (B, T) array of indices in the current context
+        for _ in range(max_new_tokens):
+            # crop idx to the last block_size tokens
+            idx_cond = idx[:, -block_size:]
+            # get the predictions
+            logits, loss = self.forward(idx_cond)
+            # focus only on the last time step
+            logits = logits[:, -1, :] # becomes (B, C)
+            # apply softmax to get probabilities
+            probs = F.softmax(logits, dim=-1) # (B, C)
+            # sample from the distribution
+            idx_next = torch.multinomial(probs, num_samples=1) # (B, 1)
+            # append sampled index to the running sequence
+            idx = torch.cat((idx, idx_next), dim=1) # (B, T+1)
+        return idx        
+
+
+# Instantiate the model
+model = BigramLanguageModel()
+
+# Specify the path to the pre-trained model checkpoint
+checkpoint_path = 'checkpoint.pth'
+
+# Load the model checkpoint
+checkpoint = torch.load(checkpoint_path)
+model.load_state_dict(checkpoint['model_state_dict'])
+model.eval()
+model.to(device)
+
+
+# generate from the model
+context = torch.zeros((1, 1), dtype=torch.long, device=device)
+
+def greet(start_character, number_of_tokens):
+    context[0][0] = encode(start_character)
+    max_new_tokens = number_of_tokens
+    return decode(model.generate(context, max_new_tokens=max_new_tokens)[0].tolist())
+
+iface = gr.Interface(fn=greet, inputs=["text", "number"], outputs="text")
 iface.launch()