recompilation of AdamW w/ OneCycleLR #133898
Description
🐛 Describe the bug
i'd like to compile my optimizer but am hitting recompilation issues. i wrap my LR in a tensor, but it seems like beta1/beta2 may need similar treatment (based on type annotations, beta is (Tuple[float, float], optional)); however, wrapping the default beta values in tensors similar to the LR breaks.
a repro:
import torch
torch._logging.set_logs(recompiles_verbose=True)
param = torch.rand(2, 3, dtype=torch.float, device="cuda", requires_grad=True)
param.grad = torch.rand_like(param)
lr = torch.tensor(0.001, device="cuda")
total_steps = 10000
optimizer = torch.optim.AdamW([param], lr=lr)
scheduler = torch.optim.lr_scheduler.OneCycleLR(
optimizer, max_lr=lr, total_steps=total_steps
)
@torch.compile()
def step():
optimizer.step()
scheduler.step()
for _ in range(total_steps):
step()recompilation
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] guard 0 failures:
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] - L['self'].param_groups[0]['betas'][0] == 0.9499993141552187
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose]
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] guard 1 failures:
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] - L['self'].param_groups[0]['betas'][0] == 0.9499995610589786
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose]
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] guard 2 failures:
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] - L['self'].param_groups[0]['betas'][0] == 0.9499997530955174
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose]
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] guard 3 failures:
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] - L['self'].param_groups[0]['betas'][0] == 0.9499998902646242
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose]
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] guard 4 failures:
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] - L['self'].param_groups[0]['betas'][0] == 0.9499999725661485
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose]
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] guard 5 failures:
V0819 21:07:52.605000 139637715699520 torch/_dynamo/guards.py:1423] [__recompiles_verbose] - L['self'].param_groups[0]['betas'][0] == 0.95
Versions
PyTorch version: 2.3.1+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
OS: Ubuntu 20.04.6 LTS (x86_64)
GCC version: (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
Clang version: Could not collect
CMake version: version 3.29.4
Libc version: glibc-2.31
Python version: 3.10.12 | packaged by conda-forge | (main, Jun 23 2023, 22:40:32) [GCC 12.3.0] (64-bit runtime)
Python platform: Linux-5.15.0-1027-oracle-x86_64-with-glibc2.31
Is CUDA available: True
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration: GPU 0: NVIDIA A100-SXM4-40GB
Nvidia driver version: 550.90.07
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.8.9.7
/usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.9.7
/usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.9.7
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.9.7
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.9.7
/usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.9.7
/usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.9.7
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
Address sizes: 48 bits physical, 48 bits virtual
CPU(s): 256
On-line CPU(s) list: 0-254
Off-line CPU(s) list: 255
Thread(s) per core: 1
Core(s) per socket: 64
Socket(s): 2
NUMA node(s): 8
Vendor ID: AuthenticAMD
CPU family: 25
Model: 1
Model name: AMD EPYC 7J13 64-Core Processor
Stepping: 1
Frequency boost: enabled
CPU MHz: 2550.000
CPU max MHz: 3673.0950
CPU min MHz: 1500.0000
BogoMIPS: 4900.16
Virtualization: AMD-V
L1d cache: 2 MiB
L1i cache: 2 MiB
L2 cache: 32 MiB
L3 cache: 256 MiB
NUMA node0 CPU(s): 0-15,128-143
NUMA node1 CPU(s): 16-31,144-159
NUMA node2 CPU(s): 32-47,160-175
NUMA node3 CPU(s): 48-63,176-191
NUMA node4 CPU(s): 64-79,192-207
NUMA node5 CPU(s): 80-95,208-223
NUMA node6 CPU(s): 96-111,224-239
NUMA node7 CPU(s): 112-127,240-254
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec store bypass: Vulnerable
Vulnerability Spectre v1: Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers
Vulnerability Spectre v2: Vulnerable, IBPB: disabled, STIBP: disabled, PBRSB-eIBRS: Not affected
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 invpcid_single hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca fsrm
Versions of relevant libraries:
[pip3] numpy==1.24.4
[pip3] numpy-stl==3.1.1
[pip3] torch==2.3.1
[pip3] torchmetrics==1.3.2
[pip3] torchvision==0.18.1
[pip3] triton==2.3.1
[conda] numpy 1.24.4 pypi_0 pypi
[conda] numpy-stl 3.1.1 pypi_0 pypi
[conda] torch 2.3.1 pypi_0 pypi
[conda] torchmetrics 1.3.2 pypi_0 pypi
[conda] torchvision 0.18.1 pypi_0 pypi
[conda] triton 2.3.1 pypi_0 pypi
cc @vincentqb @jbschlosser @albanD @janeyx99 @crcrpar @ezyang @chauhang @penguinwu @mlazos