{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "wJpXpmjEYC_T"
},
"source": [
"## Building a GPT\n",
"\n",
"## Attribution\n",
"\n",
"Note that this bigram notebook is based very heavily on Andrej Karpathy's \"makemore\" aka NN-Zero-to-Hero code and videos. All credit goes to him.\n",
"\n",
"You can find his repo here: https://github.com/karpathy/ng-video-lecture as well as the makemore repo here: https://github.com/karpathy/makemore\n",
"\n",
"His video is extremely excellent and can be found here: https://www.youtube.com/watch?v=kCc8FmEb1nY\n",
"\n",
"Refer to his LICENSE file in this folder.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from IPython.display import display, HTML\n",
"display(HTML(\"\"))\n",
"display(HTML(\"\"))\n",
"display(HTML(\"\"))\n",
"\n",
"import torch\n",
"torch.set_printoptions(linewidth=230)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "h5hjCcLDr2WC",
"outputId": "ccc60f0c-fd78-4dbe-8598-0512d1036aad"
},
"outputs": [],
"source": [
"# Download the tiny shakespeare dataset\n",
"!wget https://raw.githubusercontent.com/karpathy/char-rnn/master/data/tinyshakespeare/input.txt"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "O6medjfRsLD9"
},
"outputs": [],
"source": [
"# read it in to inspect it\n",
"with open('input.txt', 'r', encoding='utf-8') as f:\n",
" text = f.read()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "6xWI_VyAsN8F",
"outputId": "ed819dd0-72e5-40a6-d2ed-928ff73bfda6"
},
"outputs": [],
"source": [
"print(\"length of dataset in characters: \", len(text))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "2c5V0FvqseE0",
"outputId": "25ca7adc-b8c0-42d1-b08c-e0863c5c314e"
},
"outputs": [],
"source": [
"# First 1000 characters\n",
"print(text[:1000])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "0e-Rbyr8sfM8",
"outputId": "f34e94a9-5b44-4cf3-885b-986731929109"
},
"outputs": [],
"source": [
"# here are all the unique characters that occur in this text\n",
"chars = sorted(list(set(text)))\n",
"vocab_size = len(chars)\n",
"print(''.join(chars))\n",
"print(vocab_size)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Yw1LKNCgwjj1",
"outputId": "86fcc21c-2cf7-40d9-cd7b-b5a253da4459"
},
"outputs": [],
"source": [
"# 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",
"print(encode(\"hii there\"))\n",
"print(decode(encode(\"hii there\")))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "YJb0OXPwzvqg",
"outputId": "db7297cc-36a9-4fae-e941-e7bb9e0e91d1"
},
"outputs": [],
"source": [
"# let's now encode the entire text dataset and store it into a torch.Tensor\n",
"import torch # we use PyTorch: https://pytorch.org\n",
"data = torch.tensor(encode(text), dtype=torch.long)\n",
"print(data.shape, data.dtype)\n",
"print(data[:1000]) # the 1000 characters we looked at earier will to the GPT look like this"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "f_WIXqxz0lU5"
},
"outputs": [],
"source": [
"# Let's now split up the data into train and validation sets\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": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "TD5Bj8Y6IAD4",
"outputId": "bf23c586-1d33-4af1-b63d-ce6f90b0a528"
},
"outputs": [],
"source": [
"block_size = 8\n",
"train_data[:block_size+1]"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "9HXDe8vGJCEn",
"outputId": "588663aa-1de5-4ef7-aba0-4a96fe828353"
},
"outputs": [],
"source": [
"x = train_data[:block_size]\n",
"y = train_data[1:block_size+1]\n",
"for t in range(block_size):\n",
" context = x[:t+1]\n",
" target = y[t]\n",
" print(f\"when input is {context} the target: {target}\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Q3k1Czf7LuA9",
"outputId": "4ea8e8a0-443c-49bb-b3bf-ba36e1712999"
},
"outputs": [],
"source": [
"torch.manual_seed(1337)\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",
"\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",
" return x, y\n",
"\n",
"xb, yb = get_batch('train')\n",
"print('inputs:')\n",
"print(xb.shape)\n",
"print(xb)\n",
"print('targets:')\n",
"print(yb.shape)\n",
"print(yb)\n",
"\n",
"print('----')\n",
"\n",
"for b in range(batch_size): # batch dimension\n",
" for t in range(block_size): # time dimension\n",
" context = xb[b, :t+1]\n",
" target = yb[b,t]\n",
" print(f\"when input is {context.tolist()} the target: {target}\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "qpyyAeIzQjlO",
"outputId": "a650f8dc-da81-400b-bc59-0a595487fdb9"
},
"outputs": [],
"source": [
"print(xb) # our input to the transformer"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "nql_1ER53oCf",
"outputId": "5de90b1b-4603-428a-f571-fe4bd3c45436"
},
"outputs": [],
"source": [
"import torch\n",
"import torch.nn as nn\n",
"from torch.nn import functional as F\n",
"torch.manual_seed(1337)\n",
"\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",
"m = BigramLanguageModel(vocab_size)\n",
"logits, loss = m(xb, yb)\n",
"print(logits.shape)\n",
"print(loss)\n",
"\n",
"print(decode(m.generate(idx = torch.zeros((1, 1), dtype=torch.long), max_new_tokens=100)[0].tolist()))\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "eTyJ8qAaDdiF"
},
"outputs": [],
"source": [
"# create a PyTorch optimizer\n",
"optimizer = torch.optim.AdamW(m.parameters(), lr=1e-3)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Hs4kI8YdEkQj",
"outputId": "42ded55c-2983-4d91-c528-675b2edfa849"
},
"outputs": [],
"source": [
"batch_size = 32\n",
"for steps in range(100): # increase number of steps for good results...\n",
"\n",
" # sample a batch of data\n",
" xb, yb = get_batch('train')\n",
"\n",
" # evaluate the loss\n",
" logits, loss = m(xb, yb)\n",
" optimizer.zero_grad(set_to_none=True)\n",
" loss.backward()\n",
" optimizer.step()\n",
"\n",
"print(loss.item())\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "EcVIDWAZEtjN",
"outputId": "0ad6f9d2-ad58-4498-a5f8-6f31407bb18b"
},
"outputs": [],
"source": [
"print(decode(m.generate(idx = torch.zeros((1, 1), dtype=torch.long), max_new_tokens=500)[0].tolist()))"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "XinV8nmAnmKN"
},
"source": [
"## The mathematical trick in self-attention"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "tukiH-NbRBhA",
"outputId": "d981f6d4-ac08-4ec2-8284-82f5fa1e0815"
},
"outputs": [],
"source": [
"# toy example illustrating how matrix multiplication can be used for a \"weighted aggregation\"\n",
"torch.manual_seed(42)\n",
"a = torch.tril(torch.ones(3, 3))\n",
"a = a / torch.sum(a, 1, keepdim=True)\n",
"b = torch.randint(0,10,(3,2)).float()\n",
"c = a @ b\n",
"print('a=')\n",
"print(a)\n",
"print('--')\n",
"print('b=')\n",
"print(b)\n",
"print('--')\n",
"print('c=')\n",
"print(c)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Hs_E24uRE8kr",
"outputId": "8bf3ff5f-565e-48b8-de8e-7272706c8e12"
},
"outputs": [],
"source": [
"# consider the following toy example:\n",
"\n",
"torch.manual_seed(1337)\n",
"B,T,C = 4,8,2 # batch, time, channels\n",
"x = torch.randn(B,T,C)\n",
"x.shape"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "86NuXX0fn7ps"
},
"outputs": [],
"source": [
"# We want x[b,t] = mean_{i<=t} x[b,i]\n",
"xbow = torch.zeros((B,T,C))\n",
"for b in range(B):\n",
" for t in range(T):\n",
" xprev = x[b,:t+1] # (t,C)\n",
" xbow[b,t] = torch.mean(xprev, 0)\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "yhdOAd6-wXkZ",
"outputId": "eaf6ab61-dff1-4bb7-e623-47f692bad5f9"
},
"outputs": [],
"source": [
"# version 2: using matrix multiply for a weighted aggregation\n",
"wei = torch.tril(torch.ones(T, T))\n",
"wei = wei / wei.sum(1, keepdim=True)\n",
"xbow2 = wei @ x # (B, T, T) @ (B, T, C) ----> (B, T, C)\n",
"torch.allclose(xbow, xbow2)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "wOURrfG-ysoL",
"outputId": "080b500d-8110-4602-fcef-7d6f2ebfc6bc"
},
"outputs": [],
"source": [
"# version 3: use Softmax\n",
"tril = torch.tril(torch.ones(T, T))\n",
"wei = torch.zeros((T,T))\n",
"wei = wei.masked_fill(tril == 0, float('-inf'))\n",
"wei = F.softmax(wei, dim=-1)\n",
"xbow3 = wei @ x\n",
"torch.allclose(xbow, xbow3)\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "EDarxEWIRMKq",
"outputId": "07b587dd-a91c-4bb0-d7f1-e247cd5dacb5"
},
"outputs": [],
"source": [
"# version 4: self-attention!\n",
"torch.manual_seed(1337)\n",
"B,T,C = 4,8,32 # batch, time, channels\n",
"x = torch.randn(B,T,C)\n",
"\n",
"if 0:\n",
" # let's see a single Head perform self-attention\n",
" head_size = 16\n",
" key = nn.Linear(C, head_size, bias=False)\n",
" query = nn.Linear(C, head_size, bias=False)\n",
" value = nn.Linear(C, head_size, bias=False)\n",
" k = key(x) # (B, T, 16)\n",
" q = query(x) # (B, T, 16)\n",
" wei = q @ k.transpose(-2, -1) # (B, T, 16) @ (B, 16, T) ---> (B, T, T)\n",
"else:\n",
" wei = torch.zeros((T,T))\n",
"\n",
"tril = torch.tril(torch.ones(T, T))\n",
"\n",
"wei = wei.masked_fill(tril == 0, float('-inf'))\n",
"wei = F.softmax(wei, dim=-1)\n",
"\n",
"v = value(x)\n",
"out = wei @ v\n",
"#out = wei @ x\n",
"\n",
"out.shape"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "vT1hdtzXCjgL",
"outputId": "6d2c569b-7922-451f-9934-0fc564678d17"
},
"outputs": [],
"source": [
"wei"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "M5CvobiQ0pLr"
},
"source": [
"Notes:\n",
"- Attention is a **communication mechanism**. Can be seen as nodes in a directed graph looking at each other and aggregating information with a weighted sum from all nodes that point to them, with data-dependent weights.\n",
"- There is no notion of space. Attention simply acts over a set of vectors. This is why we need to positionally encode tokens.\n",
"- Each example across batch dimension is of course processed completely independently and never \"talk\" to each other\n",
"- In an \"encoder\" attention block just delete the single line that does masking with `tril`, allowing all tokens to communicate. This block here is called a \"decoder\" attention block because it has triangular masking, and is usually used in autoregressive settings, like language modeling.\n",
"- \"self-attention\" just means that the keys and values are produced from the same source as queries. In \"cross-attention\", the queries still get produced from x, but the keys and values come from some other, external source (e.g. an encoder module)\n",
"- \"Scaled\" attention additional divides `wei` by 1/sqrt(head_size). This makes it so when input Q,K are unit variance, wei will be unit variance too and Softmax will stay diffuse and not saturate too much. Illustration below"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "4SNbLq5z3oBw"
},
"outputs": [],
"source": [
"k = torch.randn(B,T,head_size)\n",
"q = torch.randn(B,T,head_size)\n",
"wei = q @ k.transpose(-2, -1) * head_size**-0.5"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Nl6I9n9IRTSo",
"outputId": "0c5b9cd0-af8a-4564-fbad-41d844e54822"
},
"outputs": [],
"source": [
"k.var()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "T1tQx7oeRvtc",
"outputId": "3541ca1a-7447-4ef7-835e-81824aebc1b5"
},
"outputs": [],
"source": [
"q.var()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "MLb_odHU3iKM",
"outputId": "a687a222-5a2c-4cdb-c1bf-17cd05b45b69"
},
"outputs": [],
"source": [
"wei.var()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "JB82yzt44REI",
"outputId": "f07da2f1-10bb-4a7a-bcaa-578587977d00"
},
"outputs": [],
"source": [
"torch.softmax(torch.tensor([0.1, -0.2, 0.3, -0.2, 0.5]), dim=-1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Mpt8569BB9_f",
"outputId": "5d8b910a-6192-44ba-ebb2-497d88e0b629"
},
"outputs": [],
"source": [
"torch.softmax(torch.tensor([0.1, -0.2, 0.3, -0.2, 0.5])*8, dim=-1) # gets too peaky, converges to one-hot"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "2Num7sX9CKOH",
"outputId": "929ceb78-a639-41d6-aac7-12997b5c93f0"
},
"outputs": [],
"source": [
"class LayerNorm1d: # (used to be BatchNorm1d)\n",
"\n",
" def __init__(self, dim, eps=1e-5, momentum=0.1):\n",
" self.eps = eps\n",
" self.gamma = torch.ones(dim)\n",
" self.beta = torch.zeros(dim)\n",
"\n",
" def __call__(self, x):\n",
" # calculate the forward pass\n",
" xmean = x.mean(1, keepdim=True) # batch mean\n",
" xvar = x.var(1, keepdim=True) # batch variance\n",
" xhat = (x - xmean) / torch.sqrt(xvar + self.eps) # normalize to unit variance\n",
" self.out = self.gamma * xhat + self.beta\n",
" return self.out\n",
"\n",
" def parameters(self):\n",
" return [self.gamma, self.beta]\n",
"\n",
"torch.manual_seed(1337)\n",
"module = LayerNorm1d(100)\n",
"x = torch.randn(32, 100) # batch size 32 of 100-dimensional vectors\n",
"x = module(x)\n",
"x.shape"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "633T2cmnW1uk",
"outputId": "7720fa58-0478-4e8a-86a7-502d4cce9443"
},
"outputs": [],
"source": [
"x[:,0].mean(), x[:,0].std() # mean,std of one feature across all batch inputs"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "LN9cK9BoXCYb",
"outputId": "6368ece0-600e-417d-8a91-7c1e5d750ba8"
},
"outputs": [],
"source": [
"x[0,:].mean(), x[0,:].std() # mean,std of a single input from the batch, of its features"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "dRJH6wM_XFfU"
},
"outputs": [],
"source": [
"# French to English translation example:\n",
"\n",
"# <--------- ENCODE ------------------><--------------- DECODE ----------------->\n",
"# les réseaux de neurones sont géniaux! neural networks are awesome!\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "ZcvKeBXoZFOY"
},
"source": [
"### Params and pre-processing"
]
},
{
"cell_type": "code",
"execution_count": null,
"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 = 16 # how many independent sequences will we process in parallel?\n",
"block_size = 32 # what is the maximum context length for predictions?\n",
"max_iters = 5000\n",
"eval_interval = 100\n",
"learning_rate = 1e-3\n",
"device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
"eval_iters = 200\n",
"n_embd = 64\n",
"n_head = 4\n",
"n_layer = 4\n",
"dropout = 0.0\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"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Self-attention head"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"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",
" B,T,C = x.shape\n",
" k = self.key(x) # (B,T,C)\n",
" q = self.query(x) # (B,T,C)\n",
" # compute attention scores (\"affinities\")\n",
" wei = q @ k.transpose(-2,-1) * C**-0.5 # (B, T, C) @ (B, C, 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,C)\n",
" out = wei @ v # (B, T, T) @ (B, T, C) -> (B, T, C)\n",
" return out"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Our new bigram model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"class BigramLanguageModel(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",
" 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"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "hoelkOrFY8bN",
"outputId": "961304cd-e379-40d4-dd56-8de0b91d2861"
},
"outputs": [],
"source": [
"\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(n_embd, 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",
"model = BigramLanguageModel()\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=2000)[0].tolist()))\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "fjjvMifYZf7x"
},
"outputs": [],
"source": []
}
],
"metadata": {
"colab": {
"provenance": []
},
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"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": "3.10.12"
}
},
"nbformat": 4,
"nbformat_minor": 1
}