Make PyTorch models Lightning fast.

---

Lightning.ai • Performance • Get started • Install • Examples • Features • Documentation •

Thunder makes PyTorch models Lightning fast.

Thunder is a source-to-source compiler for PyTorch. It makes PyTorch programs faster by combining and using different hardware executors at once (ie: nvFuser, torch.compile, cuDNN, and TransformerEngine FP8).

Works on single accelerators and in multi-GPU settings. Thunder aims to be usable, understandable, and extensible.

Thunder can achieve significant speedups over standard PyTorch eager code, through the compounding effects of optimizations and the use of best-in-class executors. Here is an example of the pretraining throughput for Llama 2 7B as implemented in LitGPT.

Thunder achieves a 40% speedup in training throughput compared to eager code on H100 using a combination of executors including nvFuser, torch.compile, cuDNN, and TransformerEngine FP8.

Thunder supports distributed strategies like DDP and FSDP (ZeRO2 and ZeRO3). Here is the normalized throughput measured for Llama 2 7B (this time without FP8 mixed precision, support for FSDP is underway).

NOTE: Lightning Thunder is alpha. Feel free to get involved, expect a few bumps along the way.

Try Thunder without installing by using our Zero to Thunder Tutorial Studio.

Install nvFuser nightly, and Thunder together

# install nvFuser which installs the matching nightly PyTorch
pip install --pre 'nvfuser-cu121[torch]' --extra-index-url https://pypi.nvidia.com

# install thunder
pip install lightning-thunder

pip install git+https://github.com/Lightning-AI/lightning-thunder.git

pip install -e .

Here is a simple example of how Thunder lets you compile and run PyTorch code:

import torch
import thunder


def foo(a, b):
    return a + b


jfoo = thunder.jit(foo)

a = torch.full((2, 2), 1)
b = torch.full((2, 2), 3)

result = jfoo(a, b)

print(result)

# prints
# tensor(
#  [[4, 4]
#   [4, 4]])

The compiled function jfoo takes and returns PyTorch tensors, just like the original function, so modules and functions compiled by Thunder can be used as part of larger PyTorch programs.

Thunder is in its early stages and should not be used for production runs yet.

However, it can already deliver outstanding performance on LLM model supported by LitGPT, such as Mistral, Llama 2, Gemma, Falcon, and others.

Check out the LitGPT integration to learn about running LitGPT and Thunder together.

Given a Python callable or PyTorch module, Thunder can generate an optimized program that:

• Computes its forward and backward passes
• Coalesces operations into efficient fusion regions
• Dispatches computations to optimized kernels
• Distributes computations optimally across machines

To do so, Thunder ships with:

• A JIT for acquiring Python programs targeting PyTorch and custom operations
• A multi-level IR to represent operations as a trace of a reduced op-set
• An extensible set of transformations on the trace, such as grad, fusions, distributed (like ddp, fsdp), functional (like vmap, vjp, jvp)
• A way to dispatch operations to an extensible collection of executors

Thunder is written entirely in Python. Even its trace is represented as valid Python at all stages of transformation. This allows unprecedented levels of introspection and extensibility.

Thunder doesn't generate code for accelerators directly. It acquires and transforms user programs so that it's possible to optimally select or generate device code using fast executors like:

• torch.compile
• nvFuser
• cuDNN
• Apex
• TransformerEngine
• PyTorch eager
• custom kernels, including those written with OpenAI Triton

Modules and functions compiled with Thunder fully interoperate with vanilla PyTorch and support PyTorch's autograd. Also, Thunder works alongside torch.compile to leverage its state-of-the-art optimizations.

Docs are currently not hosted publicly. However you can build them locally really quickly:

make docs

and point your browser to the generated docs at docs/build/index.html.

You can set up your environment for developing Thunder by installing the development requirements:

pip install -r requirements/devel.txt

Install Thunder as an editable package (optional):

pip install -e .

Now you run tests:

pytest thunder/tests

Thunder is very thoroughly tested, so expect this to take a while.

Lightning Thunder is released under the Apache 2.0 license. See the LICENSE file for details.