model file updated

transformer.py

@@ -1,53 +1,37 @@
-# Solving for residual std scaling issue
-import os
 import math
-import time
-import inspect
-from dataclasses import dataclass
 import torch
 import torch.nn as nn
 from torch.nn import functional as F
-
+from dataclasses import dataclass
 
 class CausalSelfAttention(nn.Module):
-
     def __init__(self, config):
         super().__init__()
         assert config.n_embd % config.n_head == 0
-        # key, query, value projections for all heads, but in a batch
         self.c_attn = nn.Linear(config.n_embd, 3 * config.n_embd)
-        # output projection
         self.c_proj = nn.Linear(config.n_embd, config.n_embd)
         self.c_proj.NANGPT_SCALE_INIT = 1
-        # regularization
         self.n_head = config.n_head
         self.n_embd = config.n_embd
         self.register_buffer("bias", torch.tril(torch.ones(config.block_size, config.block_size)).view(1, 1, config.block_size, config.block_size))
 
     def forward(self, x):
-        B, T, C = x.size() # batch size, sequence length, embedding dimensionality (n_embd)
-        # calculate query, key, values for all heads in batch and move head forward to be the batch dim
-        # nh is "number of heads", hs is "head size", and C (number of channels) = nh * hs
-        # e.g. in GPT-2 (124M), n_head=12, hs=64, so nh*hs=C=768 channels in the Transformer
+        B, T, C = x.size()
         qkv = self.c_attn(x)
         q, k, v = qkv.split(self.n_embd, dim=2)
-        k = k.view(B, T, self.n_head, C // self.n_head).transpose(1, 2) # (B, nh, T, hs)
-        q = q.view(B, T, self.n_head, C // self.n_head).transpose(1, 2) # (B, nh, T, hs)
-        v = v.view(B, T, self.n_head, C // self.n_head).transpose(1, 2) # (B, nh, T, hs)
+        k = k.view(B, T, self.n_head, C // self.n_head).transpose(1, 2)
+        q = q.view(B, T, self.n_head, C // self.n_head).transpose(1, 2)
+        v = v.view(B, T, self.n_head, C // self.n_head).transpose(1, 2)
 
         att = (q @ k.transpose(-2, -1)) * (1.0 / math.sqrt(k.size(-1)))
         att = att.masked_fill(self.bias[:, :, :T, :T] == 0, float('-inf'))
         att = F.softmax(att, dim=-1)
-        y = att @ v # (B, nh, T, T) x (B, nh, T, hs) -> (B, nh, T, hs)
-
-        y = y.transpose(1, 2).contiguous().view(B, T, C) # re-assemble all head outputs side by side
-        # output projection
+        y = att @ v
+        y = y.transpose(1, 2).contiguous().view(B, T, C)
         y = self.c_proj(y)
         return y
 
-
 class MLP(nn.Module):
-
     def __init__(self, config):
         super().__init__()
         self.c_fc    = nn.Linear(config.n_embd, 4 * config.n_embd)
@@ -62,7 +46,6 @@ class MLP(nn.Module):
         return x
 
 class Block(nn.Module):
-
     def __init__(self, config):
         super().__init__()
         self.ln_1 = nn.LayerNorm(config.n_embd)
@@ -75,18 +58,15 @@ class Block(nn.Module):
         x = x + self.mlp(self.ln_2(x))
         return x
 
-
 @dataclass
 class GPTConfig:
-    block_size: int = 1024 # max sequence length
-    vocab_size: int = 50257 # number of tokens: 50,000 BPE merges + 256 bytes tokens + 1 <|endoftext|> token
-    n_layer: int = 12 # number of layers
-    n_head: int = 12 # number of heads
-    n_embd: int = 768 # embedding dimension
-
+    block_size: int = 1024
+    vocab_size: int = 50257
+    n_layer: int = 12
+    n_head: int = 12
+    n_embd: int = 768
 
 class GPT(nn.Module):
-
     def __init__(self, config):
         super().__init__()
         self.config = config
@@ -98,11 +78,7 @@ class GPT(nn.Module):
             ln_f = nn.LayerNorm(config.n_embd),
         ))
         self.lm_head = nn.Linear(config.n_embd, config.vocab_size, bias=False)
-
-        # weight sharing
         self.transformer.wte.weight = self.lm_head.weight
-
-        # weight initialization
         self.apply(self._init_weights)
 
     def _init_weights(self, module):
@@ -116,218 +92,22 @@ class GPT(nn.Module):
         elif isinstance(module, nn.Embedding):
             torch.nn.init.normal_(module.weight, mean=0.0, std = 0.02)
 
-
-
     def forward(self, idx, targets=None):
-        # idx is of shape (B, T)
         B, T = idx.size()
         assert T <= self.config.block_size, f"Cannot forward sequence of length {T}, block size is only {self.config.block_size}"
-        # forward the token and posisition embeddings
-        pos = torch.arange(0, T, dtype=torch.long, device=idx.device) # shape (T)
-        pos_emb = self.transformer.wpe(pos) # position embeddings of shape (T, n_embd)
-        tok_emb = self.transformer.wte(idx) # token embeddings of shape (B, T, n_embd)
+        
+        pos = torch.arange(0, T, dtype=torch.long, device=idx.device)
+        pos_emb = self.transformer.wpe(pos)
+        tok_emb = self.transformer.wte(idx)
         x = tok_emb + pos_emb
-        # forward the blocks of the transformer
+        
         for block in self.transformer.h:
             x = block(x)
-        # forward the final layernorm and the classifier
+            
         x = self.transformer.ln_f(x)
-        logits = self.lm_head(x) # (B, T, vocab_size)
+        logits = self.lm_head(x)
+        
         loss = None
         if targets is not None:
             loss = F.cross_entropy(logits.view(-1, logits.size(-1)), targets.view(-1))
-        return logits, loss
-
-    @classmethod
-    def from_pretrained(cls, model_type):
-        """Loads pretrained GPT-2 model weights from huggingface"""
-        assert model_type in {'gpt2', 'gpt2-medium', 'gpt2-large', 'gpt2-xl'}
-        from transformers import GPT2LMHeadModel
-        print("loading weights from pretrained gpt: %s" % model_type)
-
-        # n_layer, n_head and n_embd are determined from model_type
-        config_args = {
-            'gpt2':         dict(n_layer=12, n_head=12, n_embd=768),  # 124M params
-            'gpt2-medium':  dict(n_layer=24, n_head=16, n_embd=1024), # 350M params
-            'gpt2-large':   dict(n_layer=36, n_head=20, n_embd=1280), # 774M params
-            'gpt2-xl':      dict(n_layer=48, n_head=25, n_embd=1600), # 1558M params
-        }[model_type]
-        config_args['vocab_size'] = 50257 # always 50257 for GPT model checkpoints
-        config_args['block_size'] = 1024 # always 1024 for GPT model checkpoints
-        # create a from-scratch initialized minGPT model
-        config = GPTConfig(**config_args)
-        model = GPT(config)
-        sd = model.state_dict()
-        sd_keys = sd.keys()
-        sd_keys = [k for k in sd_keys if not k.endswith('.attn.bias')] # discard this mask / buffer, not a param
-
-        # init a huggingface/transformers model
-        model_hf = GPT2LMHeadModel.from_pretrained(model_type)
-        sd_hf = model_hf.state_dict()
-
-        # copy while ensuring all of the parameters are aligned and match in names and shapes
-        sd_keys_hf = sd_hf.keys()
-        sd_keys_hf = [k for k in sd_keys_hf if not k.endswith('.attn.masked_bias')] # ignore these, just a buffer
-        sd_keys_hf = [k for k in sd_keys_hf if not k.endswith('.attn.bias')] # same, just the mask (buffer)
-        transposed = ['attn.c_attn.weight', 'attn.c_proj.weight', 'mlp.c_fc.weight', 'mlp.c_proj.weight']
-        # basically the openai checkpoints use a "Conv1D" module, but we only want to use a vanilla Linear
-        # this means that we have to transpose these weights when we import them
-        assert len(sd_keys_hf) == len(sd_keys), f"mismatched keys: {len(sd_keys_hf)} != {len(sd_keys)}"
-        for k in sd_keys_hf:
-            if any(k.endswith(w) for w in transposed):
-                # special treatment for the Conv1D weights we need to transpose
-                assert sd_hf[k].shape[::-1] == sd[k].shape
-                with torch.no_grad():
-                    sd[k].copy_(sd_hf[k].t())
-            else:
-                # vanilla copy over the other parameters
-                assert sd_hf[k].shape == sd[k].shape
-                with torch.no_grad():
-                    sd[k].copy_(sd_hf[k])
-
-        return model
-
-# model = GPT.from_pretrained('gpt2')
-
-device = 'cpu'
-if torch.cuda.is_available():
-    device = 'cuda'
-elif hasattr(torch.backends, "mps") and torch.backends.mps.is_available():
-    device = "mps"
-print(f"using device: {device}")
-
-# SEED
-torch.manual_seed(1337)
-if torch.cuda.is_available():
-    torch.cuda.manual_seed(1337)
-
-# STOP
-num_return_sequences = 5
-max_length = 30
-
-
-
-import tiktoken
-
-class DataLoaderLite:
-    def __init__(self, B, T):
-        self.B = B
-        self.T = T
-
-        # at init load tokens from disk and store them in memory
-        with open('input.txt', 'r') as f:
-            text = f.read()
-        enc = tiktoken.get_encoding('gpt2') 
-        tokens = enc.encode(text)
-        self.tokens = torch.tensor(tokens)
-        print(f'loaded {len(self.tokens)} tokens')
-        print(f'1 epoch = {len(self.tokens) // (B * T)} batches')
-
-        # state
-        self.current_position = 0
-    
-    def next_batch(self):
-        B, T = self.B, self.T
-        buf = self.tokens[self.current_position: self.current_position + B * T + 1]
-        x = (buf[:-1]).view(B, T) # inputs
-        y = (buf[1:]).view(B, T) # targets
-        # advance the position in the tensor
-        self.current_position += B*T
-        # if loading the next batch would be out of bounds, reset
-        if self.current_position + (B * T + 1) > len(self.tokens):
-            self.current_position = 0
-        return x, y
-
-
-model = GPT(GPTConfig())
-model.to(device)
-
-# Increase batch size slightly but keep it manageable
-train_loader = DataLoaderLite(B=8, T=64)
-
-# Calculate total steps for one cycle
-total_steps = 10000
-print(f"Training for {total_steps} steps")
-
-# Initialize optimizer with more conservative parameters
-optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=0.1, betas=(0.9, 0.95))
-
-# Use OneCycleLR scheduler
-scheduler = torch.optim.lr_scheduler.OneCycleLR(
-    optimizer,
-    max_lr=3e-4,
-    total_steps=total_steps,
-    pct_start=0.1,  # Warm up for 10% of steps
-    anneal_strategy='cos',
-    cycle_momentum=False,
-    div_factor=25.0,  # Initial lr = max_lr/25
-    final_div_factor=10000.0,  # Min lr = initial_lr/10000
-)
-
-# Training loop
-best_loss = float('inf')
-step = 0
-losses = []  # Keep track of losses for monitoring
-last_time = time.time()
-interval = 2  # Print every 10 steps
-
-while step < total_steps and best_loss > 0.099999:
-    x, y = train_loader.next_batch()
-    x, y = x.to(device), y.to(device)
-    
-    optimizer.zero_grad()
-    logits, loss = model(x, y)
-    loss.backward()
-    
-    # Gradient clipping
-    torch.nn.utils.clip_grad_norm_(model.parameters(), 0.5)  # Reduced from 1.0
-    
-    optimizer.step()
-    scheduler.step()
-    
-    # Update best loss
-    if loss.item() < best_loss:
-        best_loss = loss.item()
-    
-    losses.append(loss.item())
-    
-    # Print progress
-    if step % interval == 0:
-        current_time = time.time()
-        time_per_batch = (current_time - last_time) / interval if step > 0 else 0
-        last_time = current_time
-        
-        # Calculate average loss over last 100 steps for stability
-        avg_loss = sum(losses[-100:]) / min(len(losses), 100)
-        
-        print(f'step {step}, '
-              f'loss: {loss.item():.4f}, '
-              f'avg_loss: {avg_loss:.4f}, '
-              f'best_loss: {best_loss:.4f}, '
-              f'lr: {scheduler.get_last_lr()[0]:.2e}, '
-              f'time/batch: {time_per_batch:.3f}s')
-    
-    step += 1
-
-print(f'Final loss: {loss.item():.6f}')
-print(f'Best loss: {best_loss:.6f}')
-print(f'Average of last 100 losses: {sum(losses[-100:]) / min(len(losses), 100):.6f}')
-
-# Save the trained model
-save_path = 'trained_model.pt'
-torch.save({
-    'model_state_dict': model.state_dict(),
-    'best_loss': best_loss,
-    'config': model.config,
-}, save_path)
-print(f"Model saved to {save_path}")
-
-# Generation code
-enc = tiktoken.get_encoding('gpt2')
-prompt = "We are accounted poor citizens, the"
-tokens = enc.encode(prompt)
-tokens = torch.tensor(tokens, dtype=torch.long)
-tokens = tokens.unsqueeze(0).repeat(num_return_sequences, 1)
-x = tokens.to(device)
-
-# Rest of generation code remains same...
+        return logits, loss