复现Andrew Karpathy的nano GPT遇CUDA设备断言触发错误求思路

我在复现Andrew Karpathy的NanoGPT项目，确认代码和原代码完全一致，无语法错误，但运行时触发RuntimeError: CUDA error: device-side assert triggered错误。以下是完整代码，求排查思路：

import torch
import torch.nn as nn
from torch.nn import functional as F

batch_size =64
block_size =256
max_iters =5000
eval_interval =500
learn_rate=3e-4
device ='cuda' if torch.cuda.is_available() else 'cpu'
eval_iters =200
n_embd=384
n_head =6
n_layer=6
dropout =0.2

print('cuda is ok' if torch.cuda.is_available() else 'cuda i`nt ok')
#-------

torch.manual_seed(1337)
with open('input.txt','r',encoding='utf-8') as f:
    text =f.read()
chars =sorted(list(set(text)))
vocab_size =len(chars)
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]
decode = lambda l: ''.join([itos[i] for i in l])
data =torch.tensor(encode(text),dtype=torch.long)
n=int(0.9*len(data))

train_data=data[:n]
val_data=data[n:]

def get_batch(split):
    data=train_data if split=='train' else val_data
    ix =torch.randint(len(data)-block_size,(batch_size,))
    x=torch.stack([data[i:i+block_size] for i in ix])
    y=torch.stack([data[i+1:block_size+i+1] for i in ix])
    x,y=x.to(device),y.to(device)
    return x,y

@torch.no_grad()
def estimate_loss():
    out={}
    model.eval()
    for split in ['train','val']:
        losses =torch.zeros(eval_iters)
        for k in range(eval_iters):
            X,Y=get_batch(split)
            logits,loss =model(X,Y)
            losses[k]=loss.item()
        out[split]=losses.mean()
    model.train()
    return out

class Head(nn.Module):
    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)))
        self.dropout =nn.Dropout(dropout)

    def forward(self,x):
        B,T,C=x.shape
        k=self.key(x)
        q=self.query(x)
        wei =q@ k.transpose(-2,-1)*k.shape[-1]**-0.5
        wei=wei.masked_fill(self.tril[:T,:T]==0,float('-inf'))
        wei=F.softmax(wei,dim=-1)
        wei =self.dropout(wei)
        v=self.value(x)
        out=wei@v
        return out

class MutiHeadAttention(nn.Module):
    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(head_size*num_heads,n_embd)
        self.dropout=nn.Dropout(dropout)
    def forward(self,x):
        out=torch.cat([h(x) for h in self.heads],dim=-1)
        out=self.dropout(self.proj(out))
        return out

class FeedForward(nn.Module):
    def __init__(self,n_embd):
        super().__init__()
        self.net=nn.Sequential(
            nn.Linear(n_embd,4*n_embd),
            nn.ReLU(),
            nn.Linear(4*n_embd,n_embd),
            nn.Dropout(dropout),
        )

    def forward(self,x):
        return  self.net(x)

class Block(nn.Module):
    def __init__(self,n_embd,n_head):
        super().__init__()
        head_size=n_embd//n_head
        self.sa =MutiHeadAttention(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.ffwd(self.ln2(x))
        return x

class GPTlanguageModel(nn.Module):

    def __init__(self):
        super().__init__()
        self.token_embedding_table=nn.Embedding(vocab_size,n_embd)
        self.position_embedding_table =nn.Embedding(vocab_size,n_embd)
        self.blocks=nn.Sequential(*[Block(n_embd,n_head=n_head) for _ in range(n_layer)])
        self.ln_f=nn.LayerNorm(n_embd)
        self.lm_head=nn.Linear(n_embd,vocab_size)
        self.apply(self._init_weights)

    def _init_weights(self,module):
        if isinstance(module,nn.Linear):
            torch.nn.init.normal_(module.weight,mean=0.0,std=0.02)
            if module.bias is not None:
                torch.nn.init.zeros_(module.bias)
        elif isinstance(module,nn.Embedding):
            torch.nn.init.normal_(module.weight,mean=0.0,std=0.02)

    def forward(self,idx,targets=None):
        B,T=idx.shape
        tok_emb=self.token_embedding_table(idx)
        pos_emb=self.position_embedding_table(torch.arange(T,device=device))
        x=tok_emb+pos_emb
        x=self.blocks(x)
        x=self.ln_f(x)
        logits=self.lm_head(x)

        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):
        for _ in range(max_new_tokens):
            idx_cond =idx[:,-block_size:]
            logits,loss=self(idx_cond)
            logits=logits[:-1:]
            probs =F.softmax(logits,dim=-1)
            idx_next=torch.multinomial(probs,num_samples=1)
            idx=torch.cat((idx,idx_next),dim=1)
        return idx

model=GPTlanguageModel()
m=model.to(device)
print(sum(p.numel() for p in m.parameters())/1e6,'M parameters')

optimizer= torch.optim.AdamW(model.parameters(),lr=learn_rate)

for iter in range(max_iters):
    if iter % eval_interval ==0 or iter == max_iters-1:
        losses =estimate_loss()
        min_train_loss=losses['train']
        if losses['train']<=min_train_loss:
            torch.save(m.state_dict(),'model.pt')
            min_train_loss=losses['train']
        print(f"step{iter}:train loss {losses['train']:.4f},val loss {losses['val']:.4f}")

    xb,yb =get_batch('train')
    logits,loss = model(xb,yb)
    optimizer.zero_grad(set_to_none=True)
    loss.backward()
    optimizer.step()
model.load_state_dict(torch.load('model.pt'))



#context =torch.zeros((1,1),dtype=torch.long,device=device)
#print(decode(m.generate(context,max_new_tokens=500)[0].tolist()))

排查思路 #

• 位置嵌入表维度错误：GPTlanguageModel初始化中，self.position_embedding_table = nn.Embedding(vocab_size, n_embd)的第一个参数应为block_size，而非vocab_size。位置嵌入的维度对应序列长度（即block_size），原代码会导致pos_emb与tok_emb维度不匹配，触发CUDA断言。
• 生成方法的索引错误：generate方法里logits=logits[:-1:]是错误切片，正确写法应为logits = logits[:, -1, :]，目的是取最后一个时间步的输出。错误切片会导致后续维度不匹配，在CUDA环境下触发断言。
• 切换到CPU排查：将device设为'cpu'运行代码，CPU会输出更具体的错误信息（比如维度不匹配、索引越界的具体位置），而CUDA的device-side assert仅返回模糊错误。
• 训练保存逻辑优化：训练循环中min_train_loss=losses['train']每次评估都会重置，应将初始化放在循环外（比如min_train_loss = float('inf')），否则永远触发模型保存条件，虽不直接导致CUDA错误，但属于逻辑缺陷。

内容的提问来源于stack exchange，提问作者open I