diff --git a/examples/demo.ipynb b/examples/demo.ipynb
index 53ed7db..ecbadad 100644
--- a/examples/demo.ipynb
+++ b/examples/demo.ipynb
@@ -40,7 +40,7 @@
"from tqdm import tqdm\n",
"\n",
"import torch\n",
- "from torch import nn, Tensor\n",
+ "from torch import nn, Tensor, tensor\n",
"import torch.nn.functional as F\n",
"\n",
"import unit_scaling as uu\n",
@@ -53,7 +53,7 @@
" input = torch.randn(*input_shape, requires_grad=True)\n",
" output = layer(input)\n",
" output.backward(torch.randn_like(output))\n",
- " print(f\"{type(layer).__name__}:\")\n",
+ " print(f\"# {type(layer).__name__}:\")\n",
" for k, v in {\n",
" \"output\": output.std(),\n",
" \"input.grad\": input.grad.std(),\n",
@@ -82,10 +82,11 @@
"vocab_size = 256\n",
"depth = 4\n",
"head_size = 64\n",
- "mlp_expansion = 4\n",
+ "mlp_expansion = 2\n",
"\n",
"# Training\n",
"n_steps = int(5e3)\n",
+ "warmup_steps = int(1e3)\n",
"batch_size = 16\n",
"sequence_length = 256\n",
"device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
@@ -196,18 +197,18 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "SpTransformerLayer:\n",
+ "# SpTransformerLayer:\n",
" output.std = 1.01\n",
" input.grad.std = 1.01\n",
" attn_qkv.weight.std = 0.05\n",
" attn_out.weight.std = 0.05\n",
" mlp_up.weight.std = 0.05\n",
" mlp_gate.weight.std = 0.05\n",
- " mlp_down.weight.std = 0.03\n",
- "attn_qkv.weight.grad.std = 3.92\n",
- "attn_out.weight.grad.std = 6.19\n",
- " mlp_up.weight.grad.std = 5.83\n",
- "mlp_gate.weight.grad.std = 5.97\n",
+ " mlp_down.weight.std = 0.04\n",
+ "attn_qkv.weight.grad.std = 3.87\n",
+ "attn_out.weight.grad.std = 6.12\n",
+ " mlp_up.weight.grad.std = 8.21\n",
+ "mlp_gate.weight.grad.std = 8.46\n",
"mlp_down.weight.grad.std = 11.61\n"
]
}
@@ -231,11 +232,12 @@
"\n",
"Replace:\n",
" - Modules `nn.*` with `uu.*`, for example `nn.Linear -> uu.Linear`.\n",
+ " - Note that these modules disable `bias` and `elementwise_affine` by default, in keeping with the recommended u-μP scheme.\n",
" - Any final projection(s) with `uu.LinearReadout`.\n",
" - Functional calls `F.*` with `U.*`.\n",
" - Residual patterns `x = x + fn(x)` with `U.residual_split` and `U.residual_add`.\n",
" - These require a `tau` value, to weight the residual versus skip contributions. A recommended scheme for choosing `tau` is described in the paper, and implemented in `uu.transformer_residual_scaling_rule()`.\n",
- " - Any container that grows with a \"depth\" hyperparameter with `uu.Stack` or `uu.DepthModuleList` (may not be nested)."
+ " - Any container that grows with a \"depth\" hyperparameter with `uu.DepthSequential` or `uu.DepthModuleList` (may not be nested)."
]
},
{
@@ -278,7 +280,7 @@
" def __init__(self, width: int) -> None:\n",
" super().__init__()\n",
" self.embedding = uu.Embedding(vocab_size, width)\n",
- " self.layers = uu.Stack(*(UmupTransformerLayer(width, i) for i in range(depth)))\n",
+ " self.layers = uu.DepthSequential(*(UmupTransformerLayer(width, i) for i in range(depth)))\n",
" self.final_norm = uu.LayerNorm(width)\n",
" self.projection = uu.LinearReadout(width, vocab_size)\n",
"\n",
@@ -307,19 +309,19 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "UmupTransformerLayer:\n",
+ "# UmupTransformerLayer:\n",
" output.std = 1.01\n",
" input.grad.std = 1.10\n",
- " attn_qkv.weight.std = 1.00\n",
+ " attn_qkv.weight.std = 0.99\n",
" attn_out.weight.std = 0.99\n",
" mlp_up.weight.std = 1.00\n",
" mlp_gate.weight.std = 1.00\n",
" mlp_down.weight.std = 1.00\n",
- "attn_qkv.weight.grad.std = 0.63\n",
+ "attn_qkv.weight.grad.std = 0.62\n",
"attn_out.weight.grad.std = 1.08\n",
- " mlp_up.weight.grad.std = 0.50\n",
- "mlp_gate.weight.grad.std = 0.51\n",
- "mlp_down.weight.grad.std = 1.00\n"
+ " mlp_up.weight.grad.std = 0.71\n",
+ "mlp_gate.weight.grad.std = 0.73\n",
+ "mlp_down.weight.grad.std = 1.01\n"
]
}
],
@@ -335,7 +337,7 @@
"\n",
"Training a u-μP model is much like training an SP model, using `uu.optim.*` in place of `torch.optim.*`.\n",
"\n",
- "u-μP optimizers are responsible for varying the learning rate based on categories of parameters in the model and the total depth of any depth-scaled layers. These categories are stored in `parameter.mup_type` and are automatically assigned by `uu.*` modules. Total depth is stored in `parameter.mup_scaling_depth` and assigned by `uu.Stack`."
+ "u-μP optimizers are responsible for varying the learning rate based on categories of parameters in the model and the total depth of any depth-scaled layers. These categories are stored in `parameter.mup_type` and are automatically assigned by `uu.*` modules. Total depth is stored in `parameter.mup_scaling_depth` and assigned by `uu.DepthSequential`."
]
},
{
@@ -348,55 +350,55 @@
"text/html": [
"\n",
- "
\n",
+ "\n",
" \n",
" \n",
- " parameter | \n",
- " mup_type | \n",
- " mup_scaling_depth | \n",
+ " parameter | \n",
+ " mup_type | \n",
+ " mup_scaling_depth | \n",
"
\n",
" \n",
" \n",
" \n",
- " embedding.weight | \n",
- " weight | \n",
- " None | \n",
+ " embedding.weight | \n",
+ " weight | \n",
+ " None | \n",
"
\n",
" \n",
- " layers.0.attn_qkv.weight | \n",
- " weight | \n",
- " 4 | \n",
+ " layers.0.attn_qkv.weight | \n",
+ " weight | \n",
+ " 4 | \n",
"
\n",
" \n",
- " layers.0.attn_out.weight | \n",
- " weight | \n",
- " 4 | \n",
+ " layers.0.attn_out.weight | \n",
+ " weight | \n",
+ " 4 | \n",
"
\n",
" \n",
- " layers.0.mlp_up.weight | \n",
- " weight | \n",
- " 4 | \n",
+ " layers.0.mlp_up.weight | \n",
+ " weight | \n",
+ " 4 | \n",
"
\n",
" \n",
- " layers.0.mlp_gate.weight | \n",
- " weight | \n",
- " 4 | \n",
+ " layers.0.mlp_gate.weight | \n",
+ " weight | \n",
+ " 4 | \n",
"
\n",
" \n",
- " layers.0.mlp_down.weight | \n",
- " weight | \n",
- " 4 | \n",
+ " layers.0.mlp_down.weight | \n",
+ " weight | \n",
+ " 4 | \n",
"
\n",
" \n",
- " projection.weight | \n",
- " output | \n",
- " None | \n",
+ " projection.weight | \n",
+ " output | \n",
+ " None | \n",
"
\n",
" \n",
"
\n"
],
"text/plain": [
- ""
+ ""
]
},
"execution_count": 8,
@@ -423,60 +425,79 @@
},
{
"cell_type": "code",
- "execution_count": 9,
+ "execution_count": 10,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
- "SP, width=128, lr=2^-14: 0it [00:00, ?it/s]W0723 23:55:07.470000 140091692951360 torch/_logging/_internal.py:1013] [2/0] Profiler function will be ignored\n",
- "SP, width=128, lr=2^-14: 5000it [00:47, 106.14it/s, loss = 2.37]\n",
- "SP, width=128, lr=2^-13: 5000it [00:42, 116.37it/s, loss = 2.02]\n",
- "SP, width=128, lr=2^-12: 5000it [00:45, 109.68it/s, loss = 1.75]\n",
- "SP, width=128, lr=2^-11: 5000it [00:44, 111.73it/s, loss = 1.67]\n",
- "SP, width=128, lr=2^-10: 5000it [00:42, 117.15it/s, loss = 1.63]\n",
- "SP, width=128, lr=2^-9: 5000it [00:42, 118.23it/s, loss = 1.61]\n",
- "SP, width=128, lr=2^-8: 5000it [00:41, 119.55it/s, loss = 2.36]\n",
- "SP, width=512, lr=2^-14: 5000it [02:33, 32.61it/s, loss = 1.62]\n",
- "SP, width=512, lr=2^-13: 5000it [02:30, 33.25it/s, loss = 1.52]\n",
- "SP, width=512, lr=2^-12: 5000it [02:30, 33.33it/s, loss = 1.36]\n",
- "SP, width=512, lr=2^-11: 5000it [02:30, 33.29it/s, loss = 1.36]\n",
- "SP, width=512, lr=2^-10: 5000it [02:30, 33.26it/s, loss = 2.38]\n",
- "SP, width=512, lr=2^-9: 5000it [02:30, 33.32it/s, loss = 2.44]\n",
- "SP, width=512, lr=2^-8: 5000it [02:30, 33.15it/s, loss = 2.46]\n",
- "u-μP, width=128, lr=2^-4: 5000it [00:51, 96.44it/s, loss = 1.77] \n",
- "u-μP, width=128, lr=2^-3: 5000it [00:50, 98.56it/s, loss = 1.74] \n",
- "u-μP, width=128, lr=2^-2: 5000it [00:50, 99.63it/s, loss = 1.65] \n",
- "u-μP, width=128, lr=2^-1: 5000it [00:52, 95.34it/s, loss = 1.74] \n",
- "u-μP, width=128, lr=2^0: 5000it [00:50, 98.17it/s, loss = 1.56] \n",
- "u-μP, width=128, lr=2^1: 5000it [00:49, 100.44it/s, loss = 1.67]\n",
- "u-μP, width=128, lr=2^2: 5000it [00:50, 99.44it/s, loss = 2.53] \n",
- "u-μP, width=512, lr=2^-4: 5000it [02:46, 30.08it/s, loss = 1.75]\n",
- "u-μP, width=512, lr=2^-3: 5000it [02:44, 30.35it/s, loss = 1.54]\n",
- "u-μP, width=512, lr=2^-2: 5000it [02:42, 30.73it/s, loss = 1.45]\n",
- "u-μP, width=512, lr=2^-1: 5000it [02:42, 30.72it/s, loss = 1.42]\n",
- "u-μP, width=512, lr=2^0: 5000it [02:42, 30.75it/s, loss = 1.39]\n",
- "u-μP, width=512, lr=2^1: 5000it [02:42, 30.72it/s, loss = 1.56]\n",
- "u-μP, width=512, lr=2^2: 5000it [02:42, 30.70it/s, loss = 2.50]\n"
+ " SP, width=128, lr=2^-13 : 0it [00:00, ?it/s]"
+ ]
+ },
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ " SP, width=128, lr=2^-13 : 5000it [00:31, 159.33it/s, loss = 2.38]\n",
+ " SP, width=128, lr=2^-12 : 5000it [00:31, 159.29it/s, loss = 2.24]\n",
+ " SP, width=128, lr=2^-11 : 5000it [00:31, 159.39it/s, loss = 1.79]\n",
+ " SP, width=128, lr=2^-10 : 5000it [00:31, 158.86it/s, loss = 1.67]\n",
+ " SP, width=128, lr=2^-9 : 5000it [00:31, 158.96it/s, loss = 1.51]\n",
+ " SP, width=128, lr=2^-8 : 5000it [00:31, 159.12it/s, loss = 1.49]\n",
+ " SP, width=128, lr=2^-7 : 5000it [00:31, 159.25it/s, loss = 2.43]\n",
+ " SP, width=128, lr=2^-6 : 5000it [00:31, 158.84it/s, loss = 2.44]\n",
+ " SP, width=512, lr=2^-13 : 5000it [01:48, 45.89it/s, loss = 1.70]\n",
+ " SP, width=512, lr=2^-12 : 5000it [01:46, 46.82it/s, loss = 1.51]\n",
+ " SP, width=512, lr=2^-11 : 5000it [01:46, 46.82it/s, loss = 1.34]\n",
+ " SP, width=512, lr=2^-10 : 5000it [01:46, 46.84it/s, loss = 1.30]\n",
+ " SP, width=512, lr=2^-9 : 5000it [01:46, 46.85it/s, loss = 2.37]\n",
+ " SP, width=512, lr=2^-8 : 5000it [01:46, 46.82it/s, loss = 2.42]\n",
+ " SP, width=512, lr=2^-7 : 5000it [01:46, 46.80it/s, loss = 2.41]\n",
+ " SP, width=512, lr=2^-6 : 5000it [01:46, 46.85it/s, loss = 2.45]\n",
+ "u-μP, width=128, lr=2^-3 : 5000it [00:42, 118.60it/s, loss = 1.80]\n",
+ "u-μP, width=128, lr=2^-2 : 5000it [00:40, 122.86it/s, loss = 1.58]\n",
+ "u-μP, width=128, lr=2^-1 : 5000it [00:40, 122.52it/s, loss = 1.68]\n",
+ "u-μP, width=128, lr=2^0 : 5000it [00:40, 122.74it/s, loss = 1.50]\n",
+ "u-μP, width=128, lr=2^1 : 5000it [00:40, 122.22it/s, loss = 1.59]\n",
+ "u-μP, width=128, lr=2^2 : 5000it [00:40, 122.27it/s, loss = 1.77]\n",
+ "u-μP, width=128, lr=2^3 : 5000it [00:40, 122.36it/s, loss = 2.43]\n",
+ "u-μP, width=128, lr=2^4 : 5000it [00:40, 122.32it/s, loss = 2.45]\n",
+ "u-μP, width=512, lr=2^-3 : 5000it [02:01, 41.10it/s, loss = 1.69]\n",
+ "u-μP, width=512, lr=2^-2 : 5000it [02:00, 41.57it/s, loss = 1.47]\n",
+ "u-μP, width=512, lr=2^-1 : 5000it [02:00, 41.56it/s, loss = 1.46]\n",
+ "u-μP, width=512, lr=2^0 : 5000it [02:00, 41.54it/s, loss = 1.36]\n",
+ "u-μP, width=512, lr=2^1 : 5000it [02:00, 41.55it/s, loss = 1.40]\n",
+ "u-μP, width=512, lr=2^2 : 5000it [02:00, 41.56it/s, loss = 1.48]\n",
+ "u-μP, width=512, lr=2^3 : 5000it [02:00, 41.49it/s, loss = 2.45]\n",
+ "u-μP, width=512, lr=2^4 : 5000it [02:00, 41.53it/s, loss = 2.46]\n"
]
}
],
"source": [
+ "def lr_schedule(step: int) -> float:\n",
+ " if step < warmup_steps:\n",
+ " return step / warmup_steps\n",
+ " a = (step - warmup_steps) * torch.pi / (n_steps - warmup_steps)\n",
+ " return tensor(a).cos().mul(.5).add(.5)\n",
+ "\n",
+ "\n",
"def run_experiment(type_: Literal[\"SP\", \"u-μP\"], width: int, lr: float) -> List[Dict[str, Any]]:\n",
" if type_ == \"u-μP\":\n",
" model = UmupTransformerDecoder(width).to(device)\n",
- " opt = uu.optim.AdamW(model.parameters(), lr=lr)\n",
+ " opt = uu.optim.AdamW(model.parameters(), lr=tensor(lr, dtype=torch.float, device=device))\n",
"\n",
" if type_ == \"SP\":\n",
" model = SpTransformerDecoder(width).to(device)\n",
- " opt = torch.optim.AdamW(model.parameters(), lr=lr)\n",
+ " opt = torch.optim.AdamW(model.parameters(), lr=tensor(lr, dtype=torch.float, device=device))\n",
"\n",
+ " schedule = torch.optim.lr_scheduler.LambdaLR(opt, lr_schedule)\n",
" def run_step(batch: Tensor) -> Tensor:\n",
" opt.zero_grad()\n",
" loss = model.loss(batch)\n",
" loss.backward()\n",
" opt.step()\n",
+ " schedule.step()\n",
" return loss\n",
"\n",
" if compile:\n",
@@ -485,17 +506,18 @@
"\n",
" log = []\n",
" log2lr = torch.tensor(lr).log2().item()\n",
- " progress = tqdm(enumerate(batches()), desc=f\"{type_}, width={width}, lr=2^{log2lr:.0f}\")\n",
+ " progress = tqdm(enumerate(batches()), desc=f\"{type_:>4}, width={width}, lr=2^{log2lr:<5.0f}\")\n",
" for step, batch in progress:\n",
" loss = run_step(batch)\n",
" log.append(dict(step=step, loss=loss.item()))\n",
- " progress.set_postfix_str(f\"loss = {loss.item():.2f}\")\n",
+ " if (step + 1) % 100 == 0:\n",
+ " progress.set_postfix_str(f\"loss = {loss.item():.2f}\")\n",
" return pd.DataFrame.from_dict(log).assign(type=type_, width=width, lr=lr)\n",
"\n",
"\n",
"type_to_lr_range = {\n",
- " \"SP\": [2**n for n in range(-14, -8 + 1)],\n",
- " \"u-μP\": [2**n for n in range(-4, 2 + 1)],\n",
+ " \"SP\": [2**n for n in range(-13, -6 + 1)],\n",
+ " \"u-μP\": [2**n for n in range(-3, 4 + 1)],\n",
"}\n",
"df = pd.concat([\n",
" run_experiment(type_=type_, width=width, lr=lr)\n",
@@ -515,12 +537,12 @@
},
{
"cell_type": "code",
- "execution_count": 10,
+ "execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
- "image/png": "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",
+ "image/png": "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",
"text/plain": [
"