mirror of
https://github.com/frankwxu/AI4DigitalForensics.git
synced 2026-02-21 11:17:53 +00:00
updated some lectures
This commit is contained in:
Frank Xu
Binary file not shown.
Binary file not shown.
Binary file not shown.
1148
lectures/14_build_GPT_Andrej/0_pytorch_tutorial_4_GPT.ipynb
Normal file
1148
lectures/14_build_GPT_Andrej/0_pytorch_tutorial_4_GPT.ipynb
Normal file
File diff suppressed because it is too large
Load Diff
245
lectures/14_build_GPT_Andrej/bigram.ipynb
Normal file
245
lectures/14_build_GPT_Andrej/bigram.ipynb
Normal file
@@ -0,0 +1,245 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 16,
|
||||
"id": "9ad55249",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"import torch\n",
|
||||
"import torch.nn as nn\n",
|
||||
"from torch.nn import functional as F\n",
|
||||
"\n",
|
||||
"# hyperparameters\n",
|
||||
"batch_size = 4 # how many independent sequences will we process in parallel?\n",
|
||||
"block_size = 8 # what is the maximum context length for predictions?\n",
|
||||
"max_iters = 3000\n",
|
||||
"eval_interval = 300\n",
|
||||
"learning_rate = 1e-2\n",
|
||||
"device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
|
||||
"eval_iters = 200\n",
|
||||
"# ------------\n",
|
||||
"\n",
|
||||
"torch.manual_seed(1337)\n",
|
||||
"\n",
|
||||
"# wget https://raw.githubusercontent.com/karpathy/char-rnn/master/data/tinyshakespeare/input.txt\n",
|
||||
"with open('input.txt', 'r', encoding='utf-8') as f:\n",
|
||||
" text = f.read()\n",
|
||||
"\n",
|
||||
"# here are all the unique characters that occur in this text\n",
|
||||
"chars = sorted(list(set(text)))\n",
|
||||
"vocab_size = len(chars)\n",
|
||||
"# create a mapping from characters to integers\n",
|
||||
"stoi = { ch:i for i,ch in enumerate(chars) }\n",
|
||||
"itos = { i:ch for i,ch in enumerate(chars) }\n",
|
||||
"encode = lambda s: [stoi[c] for c in s] # encoder: take a string, output a list of integers\n",
|
||||
"decode = lambda l: ''.join([itos[i] for i in l]) # decoder: take a list of integers, output a string\n",
|
||||
"\n",
|
||||
"# Train and test splits\n",
|
||||
"data = torch.tensor(encode(text), dtype=torch.long)\n",
|
||||
"n = int(0.9*len(data)) # first 90% will be train, rest val\n",
|
||||
"train_data = data[:n]\n",
|
||||
"val_data = data[n:]"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 17,
|
||||
"id": "cb85a506",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# data loading\n",
|
||||
"def get_batch(split):\n",
|
||||
" # generate a small batch of data of inputs x and targets y\n",
|
||||
" data = train_data if split == 'train' else val_data\n",
|
||||
" ix = torch.randint(len(data) - block_size, (batch_size,)) # shape (batch_size,)\n",
|
||||
" x = torch.stack([data[i:i+block_size] for i in ix]) # shape (batch_size, num_tokens )\n",
|
||||
" y = torch.stack([data[i+1:i+block_size+1] for i in ix])\n",
|
||||
" x, y = x.to(device), y.to(device)\n",
|
||||
" return x, y"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 18,
|
||||
"id": "b6f102b9",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"tensor([24, 43, 58, 5, 57, 1, 46, 43], device='cuda:0')"
|
||||
]
|
||||
},
|
||||
"execution_count": 18,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"x, y= get_batch(\"train\")\n",
|
||||
"y.shape\n",
|
||||
"x[0]"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 19,
|
||||
"id": "608aa909",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"@torch.no_grad()\n",
|
||||
"def estimate_loss():\n",
|
||||
" out = {}\n",
|
||||
" model.eval()\n",
|
||||
" for split in ['train', 'val']:\n",
|
||||
" losses = torch.zeros(eval_iters)\n",
|
||||
" for k in range(eval_iters):\n",
|
||||
" X, Y = get_batch(split)\n",
|
||||
" logits, loss = model(X, Y)\n",
|
||||
" losses[k] = loss.item()\n",
|
||||
" out[split] = losses.mean()\n",
|
||||
" model.train()\n",
|
||||
" return out\n",
|
||||
"\n",
|
||||
"# super simple bigram model\n",
|
||||
"class BigramLanguageModel(nn.Module):\n",
|
||||
"\n",
|
||||
" def __init__(self, vocab_size):\n",
|
||||
" super().__init__()\n",
|
||||
" # each token directly reads off the logits for the next token from a lookup table\n",
|
||||
" self.token_embedding_table = nn.Embedding(vocab_size, vocab_size)\n",
|
||||
"\n",
|
||||
" def forward(self, idx, targets=None):\n",
|
||||
"\n",
|
||||
" # idx and targets are both (B,T) tensor of integers\n",
|
||||
" logits = self.token_embedding_table(idx) # (B,T,C)\n",
|
||||
"\n",
|
||||
" if targets is None:\n",
|
||||
" loss = None\n",
|
||||
" else:\n",
|
||||
" B, T, C = logits.shape\n",
|
||||
" logits = logits.view(B*T, C)\n",
|
||||
" targets = targets.view(B*T)\n",
|
||||
" loss = F.cross_entropy(logits, targets)\n",
|
||||
"\n",
|
||||
" return logits, loss\n",
|
||||
"\n",
|
||||
" def generate(self, idx, max_new_tokens):\n",
|
||||
" # idx is (B, T) array of indices in the current context\n",
|
||||
" for _ in range(max_new_tokens):\n",
|
||||
" # get the predictions\n",
|
||||
" logits, loss = self(idx)\n",
|
||||
" # focus only on the last time step\n",
|
||||
" logits = logits[:, -1, :] # becomes (B, C)\n",
|
||||
" # apply softmax to get probabilities\n",
|
||||
" probs = F.softmax(logits, dim=-1) # (B, C)\n",
|
||||
" # sample from the distribution\n",
|
||||
" idx_next = torch.multinomial(probs, num_samples=1) # (B, 1)\n",
|
||||
" # append sampled index to the running sequence\n",
|
||||
" idx = torch.cat((idx, idx_next), dim=1) # (B, T+1)\n",
|
||||
" return idx\n",
|
||||
"\n",
|
||||
"model = BigramLanguageModel(vocab_size)\n",
|
||||
"m = model.to(device)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 20,
|
||||
"id": "1e1fd776",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"step 0: train loss 4.6475, val loss 4.6486\n",
|
||||
"step 300: train loss 3.2369, val loss 3.2433\n",
|
||||
"step 600: train loss 2.7337, val loss 2.7694\n",
|
||||
"step 900: train loss 2.6135, val loss 2.6459\n",
|
||||
"step 1200: train loss 2.5588, val loss 2.5672\n",
|
||||
"step 1500: train loss 2.5019, val loss 2.5500\n",
|
||||
"step 1800: train loss 2.4808, val loss 2.5170\n",
|
||||
"step 2100: train loss 2.4894, val loss 2.5067\n",
|
||||
"step 2400: train loss 2.5055, val loss 2.5063\n",
|
||||
"step 2700: train loss 2.4812, val loss 2.5055\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"CExfik trid owindw OLOLENCle\n",
|
||||
"\n",
|
||||
"Hiset bube t e.\n",
|
||||
"S:\n",
|
||||
"O:3pr mealatauss:\n",
|
||||
"Wantharun qur, t.\n",
|
||||
"War dilasoate awice my.\n",
|
||||
"Wh'staromzy ou wabuts, tof isth ble mil; dilincath iree sengmin lat Hetiliovets, and Win nghirilerabousel lind te l.\n",
|
||||
"HAser ce hiry ptupr aisspllw y.\n",
|
||||
"Herin's n Boopetelives\n",
|
||||
"MPOLI swod mothakleo Windo whthCoribyo touth dourive ce higend t so mower; te\n",
|
||||
"\n",
|
||||
"AN ad nterupirf s ar irist m:\n",
|
||||
"\n",
|
||||
"Thre inleronth,\n",
|
||||
"Mad\n",
|
||||
"RD?\n",
|
||||
"\n",
|
||||
"WISo myrang, be!\n",
|
||||
"KENob&hak\n",
|
||||
"Sadsal thes ghesthinin cour ay aney Iry ts I&f my ce.\n",
|
||||
"J\n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"# create a PyTorch optimizer\n",
|
||||
"optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)\n",
|
||||
"\n",
|
||||
"for iter in range(max_iters):\n",
|
||||
"\n",
|
||||
" # every once in a while evaluate the loss on train and val sets\n",
|
||||
" if iter % eval_interval == 0:\n",
|
||||
" losses = estimate_loss()\n",
|
||||
" print(f\"step {iter}: train loss {losses['train']:.4f}, val loss {losses['val']:.4f}\")\n",
|
||||
"\n",
|
||||
" # sample a batch of data\n",
|
||||
" xb, yb = get_batch('train')\n",
|
||||
"\n",
|
||||
" # evaluate the loss\n",
|
||||
" logits, loss = model(xb, yb)\n",
|
||||
" optimizer.zero_grad(set_to_none=True)\n",
|
||||
" loss.backward()\n",
|
||||
" optimizer.step()\n",
|
||||
"\n",
|
||||
"# generate from the model\n",
|
||||
"context = torch.zeros((1, 1), dtype=torch.long, device=device)\n",
|
||||
"print(decode(m.generate(context, max_new_tokens=500)[0].tolist()))"
|
||||
]
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 3
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.13.2"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 5
|
||||
}
|
||||
268
lectures/14_build_GPT_Andrej/gpt_dev.ipynb
Normal file
268
lectures/14_build_GPT_Andrej/gpt_dev.ipynb
Normal file
@@ -0,0 +1,268 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "71a7ed1e",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"# https://raw.githubusercontent.com/karpathy/ng-video-lecture/refs/heads/master/gpt.py\n",
|
||||
"# https://www.youtube.com/watch?v=kCc8FmEb1nY"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"id": "28cdaf16",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"import torch\n",
|
||||
"import torch.nn as nn\n",
|
||||
"from torch.nn import functional as F\n",
|
||||
"\n",
|
||||
"# hyperparameters\n",
|
||||
"batch_size = 64 # how many independent sequences will we process in parallel?\n",
|
||||
"block_size = 256 # what is the maximum context length for predictions?\n",
|
||||
"max_iters = 5000\n",
|
||||
"eval_interval = 500\n",
|
||||
"learning_rate = 3e-4\n",
|
||||
"device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
|
||||
"eval_iters = 200\n",
|
||||
"n_embd = 384\n",
|
||||
"n_head = 6\n",
|
||||
"n_layer = 6\n",
|
||||
"dropout = 0.2\n",
|
||||
"# ------------\n",
|
||||
"\n",
|
||||
"torch.manual_seed(1337)\n",
|
||||
"\n",
|
||||
"# wget https://raw.githubusercontent.com/karpathy/char-rnn/master/data/tinyshakespeare/input.txt\n",
|
||||
"with open('input.txt', 'r', encoding='utf-8') as f:\n",
|
||||
" text = f.read()\n",
|
||||
"\n",
|
||||
"# here are all the unique characters that occur in this text\n",
|
||||
"chars = sorted(list(set(text)))\n",
|
||||
"vocab_size = len(chars)\n",
|
||||
"# create a mapping from characters to integers\n",
|
||||
"stoi = { ch:i for i,ch in enumerate(chars) }\n",
|
||||
"itos = { i:ch for i,ch in enumerate(chars) }\n",
|
||||
"encode = lambda s: [stoi[c] for c in s] # encoder: take a string, output a list of integers\n",
|
||||
"decode = lambda l: ''.join([itos[i] for i in l]) # decoder: take a list of integers, output a string\n",
|
||||
"\n",
|
||||
"# Train and test splits\n",
|
||||
"data = torch.tensor(encode(text), dtype=torch.long)\n",
|
||||
"n = int(0.9*len(data)) # first 90% will be train, rest val\n",
|
||||
"train_data = data[:n]\n",
|
||||
"val_data = data[n:]\n",
|
||||
"\n",
|
||||
"# data loading\n",
|
||||
"def get_batch(split):\n",
|
||||
" # generate a small batch of data of inputs x and targets y\n",
|
||||
" data = train_data if split == 'train' else val_data\n",
|
||||
" ix = torch.randint(len(data) - block_size, (batch_size,))\n",
|
||||
" x = torch.stack([data[i:i+block_size] for i in ix])\n",
|
||||
" y = torch.stack([data[i+1:i+block_size+1] for i in ix])\n",
|
||||
" x, y = x.to(device), y.to(device)\n",
|
||||
" return x, y\n",
|
||||
"\n",
|
||||
"@torch.no_grad()\n",
|
||||
"def estimate_loss():\n",
|
||||
" out = {}\n",
|
||||
" model.eval()\n",
|
||||
" for split in ['train', 'val']:\n",
|
||||
" losses = torch.zeros(eval_iters)\n",
|
||||
" for k in range(eval_iters):\n",
|
||||
" X, Y = get_batch(split)\n",
|
||||
" logits, loss = model(X, Y)\n",
|
||||
" losses[k] = loss.item()\n",
|
||||
" out[split] = losses.mean()\n",
|
||||
" model.train()\n",
|
||||
" return out\n",
|
||||
"\n",
|
||||
"class Head(nn.Module):\n",
|
||||
" \"\"\" one head of self-attention \"\"\"\n",
|
||||
"\n",
|
||||
" def __init__(self, head_size):\n",
|
||||
" super().__init__()\n",
|
||||
" self.key = nn.Linear(n_embd, head_size, bias=False)\n",
|
||||
" self.query = nn.Linear(n_embd, head_size, bias=False)\n",
|
||||
" self.value = nn.Linear(n_embd, head_size, bias=False)\n",
|
||||
" self.register_buffer('tril', torch.tril(torch.ones(block_size, block_size)))\n",
|
||||
"\n",
|
||||
" self.dropout = nn.Dropout(dropout)\n",
|
||||
"\n",
|
||||
" def forward(self, x):\n",
|
||||
" # input of size (batch, time-step, channels)\n",
|
||||
" # output of size (batch, time-step, head size)\n",
|
||||
" B,T,C = x.shape\n",
|
||||
" k = self.key(x) # (B,T,hs)\n",
|
||||
" q = self.query(x) # (B,T,hs)\n",
|
||||
" # compute attention scores (\"affinities\")\n",
|
||||
" wei = q @ k.transpose(-2,-1) * k.shape[-1]**-0.5 # (B, T, hs) @ (B, hs, T) -> (B, T, T)\n",
|
||||
" wei = wei.masked_fill(self.tril[:T, :T] == 0, float('-inf')) # (B, T, T)\n",
|
||||
" wei = F.softmax(wei, dim=-1) # (B, T, T)\n",
|
||||
" wei = self.dropout(wei)\n",
|
||||
" # perform the weighted aggregation of the values\n",
|
||||
" v = self.value(x) # (B,T,hs)\n",
|
||||
" out = wei @ v # (B, T, T) @ (B, T, hs) -> (B, T, hs)\n",
|
||||
" return out\n",
|
||||
"\n",
|
||||
"class MultiHeadAttention(nn.Module):\n",
|
||||
" \"\"\" multiple heads of self-attention in parallel \"\"\"\n",
|
||||
"\n",
|
||||
" def __init__(self, num_heads, head_size):\n",
|
||||
" super().__init__()\n",
|
||||
" self.heads = nn.ModuleList([Head(head_size) for _ in range(num_heads)])\n",
|
||||
" self.proj = nn.Linear(head_size * num_heads, n_embd)\n",
|
||||
" self.dropout = nn.Dropout(dropout)\n",
|
||||
"\n",
|
||||
" def forward(self, x):\n",
|
||||
" out = torch.cat([h(x) for h in self.heads], dim=-1)\n",
|
||||
" out = self.dropout(self.proj(out))\n",
|
||||
" return out\n",
|
||||
"\n",
|
||||
"class FeedFoward(nn.Module):\n",
|
||||
" \"\"\" a simple linear layer followed by a non-linearity \"\"\"\n",
|
||||
"\n",
|
||||
" def __init__(self, n_embd):\n",
|
||||
" super().__init__()\n",
|
||||
" self.net = nn.Sequential(\n",
|
||||
" nn.Linear(n_embd, 4 * n_embd),\n",
|
||||
" nn.ReLU(),\n",
|
||||
" nn.Linear(4 * n_embd, n_embd),\n",
|
||||
" nn.Dropout(dropout),\n",
|
||||
" )\n",
|
||||
"\n",
|
||||
" def forward(self, x):\n",
|
||||
" return self.net(x)\n",
|
||||
"\n",
|
||||
"class Block(nn.Module):\n",
|
||||
" \"\"\" Transformer block: communication followed by computation \"\"\"\n",
|
||||
"\n",
|
||||
" def __init__(self, n_embd, n_head):\n",
|
||||
" # n_embd: embedding dimension, n_head: the number of heads we'd like\n",
|
||||
" super().__init__()\n",
|
||||
" head_size = n_embd // n_head\n",
|
||||
" self.sa = MultiHeadAttention(n_head, head_size)\n",
|
||||
" self.ffwd = FeedFoward(n_embd)\n",
|
||||
" self.ln1 = nn.LayerNorm(n_embd)\n",
|
||||
" self.ln2 = nn.LayerNorm(n_embd)\n",
|
||||
"\n",
|
||||
" def forward(self, x):\n",
|
||||
" x = x + self.sa(self.ln1(x))\n",
|
||||
" x = x + self.ffwd(self.ln2(x))\n",
|
||||
" return x\n",
|
||||
"\n",
|
||||
"class GPTLanguageModel(nn.Module):\n",
|
||||
"\n",
|
||||
" def __init__(self):\n",
|
||||
" super().__init__()\n",
|
||||
" # each token directly reads off the logits for the next token from a lookup table\n",
|
||||
" self.token_embedding_table = nn.Embedding(vocab_size, n_embd)\n",
|
||||
" self.position_embedding_table = nn.Embedding(block_size, n_embd)\n",
|
||||
" self.blocks = nn.Sequential(*[Block(n_embd, n_head=n_head) for _ in range(n_layer)])\n",
|
||||
" self.ln_f = nn.LayerNorm(n_embd) # final layer norm\n",
|
||||
" self.lm_head = nn.Linear(n_embd, vocab_size)\n",
|
||||
"\n",
|
||||
" # better init, not covered in the original GPT video, but important, will cover in followup video\n",
|
||||
" self.apply(self._init_weights)\n",
|
||||
"\n",
|
||||
" def _init_weights(self, module):\n",
|
||||
" if isinstance(module, nn.Linear):\n",
|
||||
" torch.nn.init.normal_(module.weight, mean=0.0, std=0.02)\n",
|
||||
" if module.bias is not None:\n",
|
||||
" torch.nn.init.zeros_(module.bias)\n",
|
||||
" elif isinstance(module, nn.Embedding):\n",
|
||||
" torch.nn.init.normal_(module.weight, mean=0.0, std=0.02)\n",
|
||||
"\n",
|
||||
" def forward(self, idx, targets=None):\n",
|
||||
" B, T = idx.shape\n",
|
||||
"\n",
|
||||
" # idx and targets are both (B,T) tensor of integers\n",
|
||||
" tok_emb = self.token_embedding_table(idx) # (B,T,C)\n",
|
||||
" pos_emb = self.position_embedding_table(torch.arange(T, device=device)) # (T,C)\n",
|
||||
" x = tok_emb + pos_emb # (B,T,C)\n",
|
||||
" x = self.blocks(x) # (B,T,C)\n",
|
||||
" x = self.ln_f(x) # (B,T,C)\n",
|
||||
" logits = self.lm_head(x) # (B,T,vocab_size)\n",
|
||||
"\n",
|
||||
" if targets is None:\n",
|
||||
" loss = None\n",
|
||||
" else:\n",
|
||||
" B, T, C = logits.shape\n",
|
||||
" logits = logits.view(B*T, C)\n",
|
||||
" targets = targets.view(B*T)\n",
|
||||
" loss = F.cross_entropy(logits, targets)\n",
|
||||
"\n",
|
||||
" return logits, loss\n",
|
||||
"\n",
|
||||
" def generate(self, idx, max_new_tokens):\n",
|
||||
" # idx is (B, T) array of indices in the current context\n",
|
||||
" for _ in range(max_new_tokens):\n",
|
||||
" # crop idx to the last block_size tokens\n",
|
||||
" idx_cond = idx[:, -block_size:]\n",
|
||||
" # get the predictions\n",
|
||||
" logits, loss = self(idx_cond)\n",
|
||||
" # focus only on the last time step\n",
|
||||
" logits = logits[:, -1, :] # becomes (B, C)\n",
|
||||
" # apply softmax to get probabilities\n",
|
||||
" probs = F.softmax(logits, dim=-1) # (B, C)\n",
|
||||
" # sample from the distribution\n",
|
||||
" idx_next = torch.multinomial(probs, num_samples=1) # (B, 1)\n",
|
||||
" # append sampled index to the running sequence\n",
|
||||
" idx = torch.cat((idx, idx_next), dim=1) # (B, T+1)\n",
|
||||
" return idx\n",
|
||||
"\n",
|
||||
"model = GPTLanguageModel()\n",
|
||||
"m = model.to(device)\n",
|
||||
"# print the number of parameters in the model\n",
|
||||
"print(sum(p.numel() for p in m.parameters())/1e6, 'M parameters')\n",
|
||||
"\n",
|
||||
"# create a PyTorch optimizer\n",
|
||||
"optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)\n",
|
||||
"\n",
|
||||
"for iter in range(max_iters):\n",
|
||||
"\n",
|
||||
" # every once in a while evaluate the loss on train and val sets\n",
|
||||
" if iter % eval_interval == 0 or iter == max_iters - 1:\n",
|
||||
" losses = estimate_loss()\n",
|
||||
" print(f\"step {iter}: train loss {losses['train']:.4f}, val loss {losses['val']:.4f}\")\n",
|
||||
"\n",
|
||||
" # sample a batch of data\n",
|
||||
" xb, yb = get_batch('train')\n",
|
||||
"\n",
|
||||
" # evaluate the loss\n",
|
||||
" logits, loss = model(xb, yb)\n",
|
||||
" optimizer.zero_grad(set_to_none=True)\n",
|
||||
" loss.backward()\n",
|
||||
" optimizer.step()\n",
|
||||
"\n",
|
||||
"# generate from the model\n",
|
||||
"context = torch.zeros((1, 1), dtype=torch.long, device=device)\n",
|
||||
"print(decode(m.generate(context, max_new_tokens=500)[0].tolist()))\n",
|
||||
"#open('more.txt', 'w').write(decode(m.generate(context, max_new_tokens=10000)[0].tolist()))"
|
||||
]
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 3
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.13.2"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 5
|
||||
}
|
||||
40000
lectures/14_build_GPT_Andrej/input.txt
Normal file
40000
lectures/14_build_GPT_Andrej/input.txt
Normal file
File diff suppressed because it is too large
Load Diff
Reference in New Issue
Block a user