|
Name |
|
Date |
Size |
#Lines |
LOC |
| .. | | - | - |
| api/ | H | 25-Apr-2025 | - | 2,408 | 1,727 |
| backends/ | H | 25-Apr-2025 | - | 3,441 | 2,396 |
| codegen/ | H | 25-Apr-2025 | - | 7,074 | 5,415 |
| cuda/ | H | 25-Apr-2025 | - | 180 | 135 |
| docs/ | H | 25-Apr-2025 | - | 426 | 332 |
| frontend/ | H | 25-Apr-2025 | - | 18,330 | 14,425 |
| ir/ | H | 25-Apr-2025 | - | 9,516 | 7,045 |
| mobile/ | H | 25-Apr-2025 | - | 11,365 | 8,201 |
| operator_upgraders/ | H | 25-Apr-2025 | - | 878 | 710 |
| passes/ | H | 25-Apr-2025 | - | 48,638 | 36,695 |
| python/ | H | 25-Apr-2025 | - | 12,474 | 10,710 |
| runtime/ | H | 25-Apr-2025 | - | 40,207 | 33,760 |
| serialization/ | H | 25-Apr-2025 | - | 14,882 | 12,269 |
| tensorexpr/ | H | 25-Apr-2025 | - | 44,716 | 36,229 |
| testing/ | H | 25-Apr-2025 | - | 789 | 641 |
| JIT-AUTOCAST.md | H A D | 25-Apr-2025 | 6.8 KiB | 225 | 165 |
| OVERVIEW.md | H A D | 25-Apr-2025 | 93.8 KiB | 1,551 | 1,176 |
| README.md | H A D | 25-Apr-2025 | 1.6 KiB | 31 | 26 |
| jit_log.cpp | H A D | 25-Apr-2025 | 5 KiB | 187 | 152 |
| jit_log.h | H A D | 25-Apr-2025 | 4.7 KiB | 127 | 68 |
| jit_opt_limit.cpp | H A D | 25-Apr-2025 | 2.2 KiB | 84 | 69 |
| jit_opt_limit.h | H A D | 25-Apr-2025 | 1.3 KiB | 38 | 8 |
| resource_guard.h | H A D | 25-Apr-2025 | 440 | 26 | 18 |
README.md
1# PyTorch JIT
2
3This folder contains (most of) the C++ code for the PyTorch JIT, a language
4and compiler stack for executing PyTorch models portably and efficiently. To
5learn more about the JIT from a user perspective, please consult our
6[reference documentation](https://pytorch.org/docs/stable/jit.html) and
7[tutorials](https://pytorch.org/tutorials/beginner/Intro_to_TorchScript_tutorial.html).
8
9A brief summary of the source tree:
10- [OVERVIEW.md](OVERVIEW.md): High-level technical overview of the JIT.
11- [frontend/](frontend): Taking PyTorch modules in Python and translating them into the
12 JIT IR.
13- [ir/](ir): Core IR abstractions.
14- [runtime/](runtime): Interpreter, graph execution, and JIT operators.
15- [codegen/](codegen): Generating efficient, hardware-specific code for JIT subgraphs.
16- [serialization/](serialization): Saving and loading modules.
17- [api/](api): Any user-facing C++ or Python interfaces.
18- [python/](python): Binding stuff into Python or accessing information from the Python
19 environment.
20- [testing/](testing): Utilities and helpers for testing.
21- [mobile/](mobile): Mobile-specific implementations of runtime components.
22- [passes/](passes): IR-to-IR passes, generally for optimization and lowering.
23- [generated/](generated): This folder is generated by the PyTorch build, and contains
24 bindings for native PyTorch operators into the JIT.
25
26**Refer** to each folder for more in-depth documentation.
27
28Other relevant parts of the codebase not contained here:
29- [aten/src/ATen/core](/aten/src/ATen/core): contains JIT code re-used by other elements of the
30 runtime system (eager, mobile, etc.)
31