🔳 l2hmc-qcd Example: 4D SU(3)

Author

Affiliation

Sam Foreman

ALCF

Imports

# %matplotlib inline
import matplotlib_inline
matplotlib_inline.backend_inline.set_matplotlib_formats('svg')
import os
os.environ['COLORTERM'] = 'truecolor'
import lovely_tensors as lt
lt.monkey_patch()
lt.set_config(color=False)
# automatically detect and reload local changes to modules
%load_ext autoreload
%autoreload 2
import ezpz
import numpy as np
import matplotlib.pyplot as plt
from l2hmc.utils.plot_helpers import FigAxes
import ambivalent
plt.style.use(ambivalent.STYLES['ambivalent'])
#set_plot_style()

Output:

[2025-04-30 15:42:06,938] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to mps (auto detect)

W0430 15:42:09.268000 24193 site-packages/torch/distributed/elastic/multiprocessing/redirects.py:29] NOTE: Redirects are currently not supported in Windows or MacOs.

Using device: cpu

Failed to download font: IBM Plex Sans, skipping! Failed to download font: IBM Plex Sans Condensed, skipping!

Failed to download font: IBM Plex Serif, skipping!

import ezpz
from pathlib import Path
from typing import Optional
from rich import print

import lovely_tensors as lt
import matplotlib.pyplot as plt
import numpy as np
import torch
import yaml

# from l2hmc.utils.dist import setup_torch
seed = np.random.randint(2 ** 32)
print(f"seed: {seed}")

_ = ezpz.setup_torch(seed=seed)
torch.set_default_dtype(torch.float64)

logger = ezpz.get_logger()

import l2hmc.group.su3.pytorch.group as g
from l2hmc.common import grab_tensor, print_dict
from l2hmc.configs import dict_to_list_of_overrides, get_experiment
from l2hmc.experiment.pytorch.experiment import Experiment, evaluate  # noqa

Output:

seed: 3107208906

[2025-04-30 15:42:15][I][ezpz/dist:557] Using get_torch_device_type()='mps' with backend='gloo'

[2025-04-30 15:42:15][I][ezpz/dist:873] Using device='mps' with backend='ddp' + 'gloo' for distributed training.

[2025-04-30 15:42:15][I][ezpz/dist:923] ['Sams-MacBook-Pro-2.local'][0/0]

from l2hmc.utils.plot_helpers import (  # noqa
    plot_scalar,
    plot_chains,
    plot_leapfrogs
)

def savefig(fig: plt.Figure, fname: str, outdir: os.PathLike):
    pngfile = Path(outdir).joinpath(f"pngs/{fname}.png")
    svgfile = Path(outdir).joinpath(f"svgs/{fname}.svg")
    pngfile.parent.mkdir(exist_ok=True, parents=True)
    svgfile.parent.mkdir(exist_ok=True, parents=True)
    fig.savefig(svgfile, transparent=True, bbox_inches='tight')
    fig.savefig(pngfile, transparent=True, bbox_inches='tight', dpi=300)

def plot_metrics(metrics: dict, title: Optional[str] = None, **kwargs):
    outdir = Path(f"./plots-4dSU3/{title}")
    outdir.mkdir(exist_ok=True, parents=True)
    for key, val in metrics.items():
        fig, ax = plot_metric(val, name=key, **kwargs)
        if title is not None:
            ax.set_title(title)
        console.log(f"Saving {key} to {outdir}")
        savefig(fig, f"{key}", outdir=outdir)
        plt.show()

def plot_metric(
    metric: torch.Tensor,
    name: Optional[str] = None,
    **kwargs,
):
    assert len(metric) > 0
    if isinstance(metric[0], (int, float, bool, np.floating)):
        y = np.stack(metric)
        return plot_scalar(y, ylabel=name, **kwargs)
    element_shape = metric[0].shape
    if len(element_shape) == 2:
        y = grab_tensor(torch.stack(metric))
        return plot_leapfrogs(y, ylabel=name)
    if len(element_shape) == 1:
        y = grab_tensor(torch.stack(metric))
        return plot_chains(y, ylabel=name, **kwargs)
    if len(element_shape) == 0:
        y = grab_tensor(torch.stack(metric))
        return plot_scalar(y, ylabel=name, **kwargs)
    raise ValueError

Load config + build Experiment

from rich import print

from l2hmc.configs import CONF_DIR
su3conf = Path(f"{CONF_DIR}/su3test.yaml")
with su3conf.open('r') as stream:
    conf = dict(yaml.safe_load(stream))

import json
from rich import print_json
print_json(json.dumps(conf, indent=4, sort_keys=True))
overrides = dict_to_list_of_overrides(conf)

Output:

{
  "annealing_schedule": {
    "beta_final": 6.0,
    "beta_init": 6.0
  },
  "backend": "DDP",
  "conv": "none",
  "dynamics": {
    "eps": 0.01,
    "eps_fixed": false,
    "group": "SU3",
    "latvolume": [
        4,
        4,
        4,
        4
    ],
    "merge_directions": true,
    "nchains": 8,
    "nleapfrog": 4,
    "use_separate_networks": false,
    "use_split_xnets": false,
    "verbose": true
  },
  "framework": "pytorch",
  "init_aim": false,
  "init_wandb": false,
  "learning_rate": {
    "clip_norm": 1.0,
    "lr_init": "1e-04"
  },
  "loss": {
    "aux_weight": 0.0,
    "charge_weight": 0.0,
    "plaq_weight": 0.1,
    "rmse_weight": 0.1,
    "use_mixed_loss": true
  },
  "net_weights": {
    "v": {
        "q": 1.0,
        "s": 1.0,
        "t": 1.0
    },
    "x": {
        "q": 1.0,
        "s": 0.0,
        "t": 1.0
    }
  },
  "network": {
    "activation_fn": "tanh",
    "dropout_prob": 0.0,
    "units": [
        256
    ],
    "use_batch_norm": false
  },
  "restore": false,
  "save": false,
  "steps": {
    "log": 1,
    "nepoch": 10,
    "nera": 1,
    "print": 1,
    "test": 50
  },
  "use_tb": false,
  "use_wandb": false
}

ptExpSU3 = get_experiment(overrides=[*overrides], build_networks=True)

# console.print(ptExpSU3.config)
state = ptExpSU3.trainer.dynamics.random_state(6.0)
logger.info(f"checkSU(state.x): {g.checkSU(state.x)}")
logger.info(f"checkSU(state.x): {g.checkSU(g.projectSU(state.x))}")
assert isinstance(state.x, torch.Tensor)
assert isinstance(state.beta, torch.Tensor)
assert isinstance(ptExpSU3, Experiment)

Output:

[2025-04-30 15:42:54][I][utils/dist:229:l2hmc.utils.dist] Caught MASTER_PORT:1234 from environment!

[2025-04-30 15:42:54][I][utils/dist:229:l2hmc.utils.dist] Caught MASTER_PORT:1234 from environment!

[2025-04-30 15:42:54][W][pytorch/trainer:467:l2hmc.trainers.pytorch.trainer] Using torch.float32 on cpu!

[2025-04-30 15:42:54][W][pytorch/trainer:467:l2hmc.trainers.pytorch.trainer] Using `torch.optim.Adam` optimizer

[2025-04-30 15:42:54][I][pytorch/trainer:305:l2hmc.trainers.pytorch.trainer] num_params in model: 27880456

[2025-04-30 15:42:55][W][pytorch/trainer:271:l2hmc.trainers.pytorch.trainer] logging with freq 1 for wandb.watch

[2025-04-30 15:42:55][I][ipykernel_24193/1455121896:5:ezpz.log] checkSU(state.x): (tensor[8] f64 x∈[1.374e-14, 2.051e-13] μ=5.186e-14 σ=6.363e-14 [2.328e-14, 1.850e-14, 5.587e-14, 2.051e-13, 4.692e-14, 1.374e-14, 1.985e-14, 3.163e-14], tensor[8] f64 x∈[2.012e-13, 6.500e-12] μ=1.497e-12 σ=2.096e-12 [4.096e-13, 4.134e-13, 1.774e-12, 6.500e-12, 1.446e-12, 2.012e-13, 4.689e-13, 7.596e-13])

[2025-04-30 15:42:55][I][ipykernel_24193/1455121896:6:ezpz.log] checkSU(state.x): (tensor[8] f64 x∈[2.705e-16, 2.883e-16] μ=2.798e-16 σ=5.148e-18 [2.763e-16, 2.705e-16, 2.803e-16, 2.883e-16, 2.825e-16, 2.819e-16, 2.804e-16, 2.783e-16], tensor[8] f64 x∈[8.900e-16, 9.337e-16] μ=9.098e-16 σ=1.819e-17 [8.903e-16, 9.331e-16, 9.147e-16, 8.903e-16, 9.337e-16, 9.114e-16, 8.900e-16, 9.147e-16])

# from l2hmc.utils.plot_helpers import set_plot_style
# set_plot_style()

from l2hmc.common import get_timestamp
TSTAMP = get_timestamp()
OUTPUT_DIR = Path(f"./outputs/pt4dSU3/{TSTAMP}")
HMC_DIR = OUTPUT_DIR.joinpath('hmc')
EVAL_DIR = OUTPUT_DIR.joinpath('eval')
TRAIN_DIR = OUTPUT_DIR.joinpath('train')
HMC_DIR.mkdir(exist_ok=True, parents=True)
EVAL_DIR.mkdir(exist_ok=True, parents=True)
TRAIN_DIR.mkdir(exist_ok=True, parents=True)

ptExpSU3.trainer.print_grads_and_weights()
logger.info(ptExpSU3.config)
#console.print(ptExpSU3.config)

Output:

[2025-04-30 15:43:03][I][pytorch/trainer:2000:l2hmc.trainers.pytorch.trainer] --------------------------------------------------------------------------------

[2025-04-30 15:43:03][I][pytorch/trainer:2001:l2hmc.trainers.pytorch.trainer] GRADS:

[2025-04-30 15:43:03][I][l2hmc/common:97] networks.xnet.input_layer.xlayer.weight: None None
None
networks.xnet.input_layer.xlayer.bias: None None
None
networks.xnet.input_layer.vlayer.weight: None None
None
networks.xnet.input_layer.vlayer.bias: None None
None
networks.xnet.scale.coeff: None None
None
networks.xnet.scale.layer.weight: None None
None
networks.xnet.scale.layer.bias: None None
None
networks.xnet.transf.coeff: None None
None
networks.xnet.transf.layer.weight: None None
None
networks.xnet.transf.layer.bias: None None
None
networks.xnet.transl.weight: None None
None
networks.xnet.transl.bias: None None
None
networks.vnet.input_layer.xlayer.weight: None None
None
networks.vnet.input_layer.xlayer.bias: None None
None
networks.vnet.input_layer.vlayer.weight: None None
None
networks.vnet.input_layer.vlayer.bias: None None
None
networks.vnet.scale.coeff: None None
None
networks.vnet.scale.layer.weight: None None
None
networks.vnet.scale.layer.bias: None None
None
networks.vnet.transf.coeff: None None
None
networks.vnet.transf.layer.weight: None None
None
networks.vnet.transf.layer.bias: None None
None
networks.vnet.transl.weight: None None
None
networks.vnet.transl.bias: None None
None
xeps.0: None None
None
xeps.1: None None
None
xeps.2: None None
None
xeps.3: None None
None
veps.0: None None
None
veps.1: None None
None
veps.2: None None
None
veps.3: None None
None

[2025-04-30 15:43:03][I][pytorch/trainer:2003:l2hmc.trainers.pytorch.trainer] --------------------------------------------------------------------------------

[2025-04-30 15:43:03][I][pytorch/trainer:2004:l2hmc.trainers.pytorch.trainer] WEIGHTS:

[2025-04-30 15:43:03][I][l2hmc/common:97] networks.xnet.input_layer.xlayer.weight: torch.Size([256, 18432]) torch.float64
[[-4.19951343e-03 -4.26238340e-03  2.35647156e-03 ...  1.68137878e-03
  -1.50068399e-03  3.15863944e-03]
 [ 2.39342553e-04 -4.97551288e-03  7.26050381e-03 ... -5.64728922e-03
  -5.55020874e-03  1.17323965e-03]
 [ 5.26405398e-03 -3.91491501e-03  7.14260998e-04 ... -6.31723398e-03
   5.23567487e-03  2.05607864e-03]
 ...
 [ 6.68730979e-03 -6.91921820e-03  8.95772739e-05 ... -4.07246778e-03
  -2.29316978e-03 -3.43056826e-03]
 [-1.77016491e-03 -1.95781418e-03 -5.01585640e-03 ... -6.46914302e-03
   4.20147742e-03 -7.09234809e-03]
 [-1.94973833e-03 -7.26403514e-03  4.98051500e-04 ... -6.28795095e-03
   6.79912449e-04 -9.21666063e-04]]
networks.xnet.input_layer.xlayer.bias: torch.Size([256]) torch.float64
[ 3.15653089e-03 -6.53747057e-03 -6.14933637e-03 -5.49529500e-03
 -4.48003873e-03 -7.03491801e-03 -6.56321877e-03  4.67182499e-03
  4.36532234e-03 -5.36476346e-03 -6.09211776e-03 -7.10460550e-03
 -6.52009109e-03 -6.43876671e-03 -1.97297283e-03  5.79927976e-03
 -5.49722940e-03  6.75426645e-03  6.06732313e-03  2.56460577e-03
 -5.80235106e-03  1.03618414e-03  6.36349657e-04  1.60436063e-03
 -3.06399032e-05 -7.50359147e-04  4.36385955e-03  1.00302522e-03
  7.32044498e-03 -1.43824191e-03 -2.50594005e-03 -6.82552956e-03
 -4.71028329e-03  1.71534301e-03  6.10709162e-03 -4.05667409e-03
 -2.30713007e-03 -5.12144947e-03 -7.15109109e-04 -4.44328498e-03
 -4.94695187e-04  3.15081409e-03 -2.37185464e-03 -1.91157742e-03
  6.91548413e-03  3.81991132e-04 -4.31922342e-03 -1.98542428e-03
  4.05912039e-03  1.92396383e-03  3.53810871e-03  6.53665668e-04
 -5.06483635e-04 -5.86019994e-03 -2.86685524e-03 -2.27205811e-03
  7.25025555e-03 -5.94056580e-03 -1.54673733e-03  8.73715266e-04
 -4.36832096e-03  4.80786903e-03  1.69290306e-03 -1.38375220e-03
  6.96793511e-03  5.39967604e-03 -4.38277881e-03  4.86837191e-03
 -4.63730538e-03 -2.72580171e-04  5.81442710e-03  1.45265237e-03
  5.09906755e-03 -5.65087549e-03 -1.40143350e-03 -6.84733200e-03
  1.19933276e-03  7.28487144e-03 -1.88899102e-03 -1.74111447e-03
 -2.23510673e-03  2.73357402e-03 -1.06047067e-03 -5.84367407e-04
 -1.07953771e-03 -7.04448438e-03  3.29054140e-03  1.94660675e-04
 -1.10677934e-03 -2.41696262e-03 -1.78325949e-03  1.03577164e-03
  7.27804628e-03 -5.85394151e-03 -4.49392971e-03 -6.15160667e-03
 -4.20118105e-03  4.69744100e-03 -4.00506019e-03 -3.18274408e-03
  6.46112415e-03  2.89159411e-03  5.58561882e-05 -2.93133409e-03
  8.16798907e-04 -6.32176171e-03  5.56053897e-03 -6.36041217e-04
 -5.79233323e-03  5.06384306e-03 -3.84639758e-03  4.87318855e-03
  2.09912290e-04  4.33691731e-03  3.37906617e-03 -5.05196960e-03
  1.02895130e-03  4.08134740e-03 -1.53992986e-03  2.17678400e-03
 -6.35111539e-03 -5.66248347e-03 -1.10881411e-03 -4.16939110e-03
 -2.41698897e-03  7.25206709e-03  3.52455892e-03 -2.72657199e-03
 -4.46049487e-03  6.27997800e-03  4.12822311e-03 -4.29568963e-03
  3.27822525e-03  4.74422437e-04 -7.36276516e-03  6.08754623e-04
  5.07836316e-03 -2.49084412e-03  6.68436632e-03  1.63677606e-03
 -6.40259777e-03 -3.89026602e-03  5.53970429e-04 -4.87297662e-03
  5.29615772e-03 -7.04606173e-03  3.74678711e-03  3.76877809e-03
  4.02093817e-03 -4.28408453e-04 -5.69630100e-03  7.00086373e-03
 -3.02440694e-03  2.58179140e-03  5.33379816e-03 -6.61146570e-03
  1.48293246e-03  2.37842092e-03 -4.01829177e-03  5.58703024e-03
 -5.74216662e-04  5.99784977e-03 -4.73664072e-03  2.37927320e-03
 -3.39304510e-03 -2.54775472e-03  4.55417377e-03  8.61042071e-04
  5.07867546e-03  5.38944149e-03  8.43247180e-05 -3.12868764e-03
 -1.34245397e-03  6.48412301e-03  7.15439817e-03  1.66651130e-04
  4.06732860e-03  2.83814521e-03  7.24635495e-03  6.22788597e-03
  4.20149888e-03  6.49253302e-04 -3.91114083e-03 -3.58432497e-03
 -3.86471098e-03  2.81108473e-03  1.78824730e-03  1.68233903e-03
  2.84089912e-03  6.25876716e-03 -4.43284225e-03  1.18857102e-03
 -4.23199176e-03  6.52209025e-03  3.10487973e-04  7.15578276e-03
 -2.14404584e-03  5.07581213e-03 -5.91204650e-03  6.67434357e-03
 -3.67739919e-04  4.18005262e-04 -2.95130127e-04  4.67975206e-03
  1.59355487e-03  5.93234468e-03 -2.09718803e-03  2.76864020e-03
  1.19992438e-03  5.40846449e-03  1.29281665e-03  7.07159381e-03
  1.41238172e-03 -5.80262790e-03  3.72717315e-03 -4.90456269e-03
 -6.72488463e-03 -5.56823441e-03 -5.23051378e-03 -6.76384914e-03
 -3.99397980e-03  3.39150096e-03  2.88789147e-03 -5.20359136e-04
  2.38952795e-03  1.60403547e-03 -2.61388670e-03  2.66299234e-03
 -8.18143279e-04 -2.57845668e-03  4.40694062e-03  2.04237922e-03
  5.15184399e-03 -3.10021599e-03  6.17029479e-03  7.23334944e-03
  3.18039373e-03 -1.86054513e-03 -3.84445095e-03  3.61909527e-03
 -6.88008789e-03 -4.94012773e-03  3.70646320e-03  5.90539612e-03
 -6.96513871e-03  5.62814758e-03 -7.02268337e-03 -7.26117535e-03
  4.51210770e-03 -4.71571024e-03 -1.22832005e-04 -8.83315043e-04
 -9.59028740e-04  6.38134499e-03 -2.59574649e-03 -1.15310570e-03]
networks.xnet.input_layer.vlayer.weight: torch.Size([256, 18432]) torch.float64
[[ 0.00463266 -0.00236656 -0.00706409 ... -0.00116637 -0.0052418
  -0.00603106]
 [-0.0070495  -0.00672454 -0.00573184 ... -0.0050628   0.00602185
  -0.00735768]
 [ 0.00696305 -0.00561956 -0.00690782 ...  0.0012015   0.00509584
   0.00545731]
 ...
 [-0.00098264  0.00689178 -0.00444889 ... -0.0024313   0.0057689
   0.00241326]
 [-0.00321381 -0.00520019 -0.00651976 ... -0.00382101 -0.00310216
  -0.00372552]
 [-0.00202929  0.00629375  0.00024414 ... -0.00602879  0.0022714
  -0.00331077]]
networks.xnet.input_layer.vlayer.bias: torch.Size([256]) torch.float64
[ 7.13954968e-03  2.33497538e-03 -1.57597373e-03 -2.46922688e-03
  1.36409845e-03 -3.54782344e-03  6.88996067e-03  1.10860886e-03
  1.38812383e-03 -7.19016211e-03  1.45510027e-03  3.02318547e-03
 -7.32138405e-03 -8.43571075e-04  4.28112049e-03  2.40885560e-03
  7.18076241e-03  1.56188999e-03  7.62330334e-04  6.74154006e-04
  3.15126920e-03  5.91587112e-03 -4.05739295e-03 -5.78828923e-03
  1.94070617e-04 -4.62861933e-03 -3.08415149e-04  3.08347898e-04
 -1.03532907e-03  1.76321011e-03  1.00590336e-03  5.19000988e-03
 -6.80118548e-03 -3.83583844e-03  2.23868496e-03 -3.56978474e-03
  4.79782172e-03 -2.77326189e-03  7.04744644e-03  3.53315120e-03
 -3.15971697e-03 -3.91518415e-03 -1.13455609e-03 -9.63908149e-04
  2.64954937e-03  5.46426969e-03  4.37747118e-03  3.15638766e-03
 -1.92890249e-03  5.45232414e-03  3.35068143e-03  4.33478276e-04
 -4.31518024e-03 -6.72472428e-03  4.54341573e-03 -2.88636991e-03
  3.70153548e-03 -6.86245384e-03  4.75460380e-03  6.41603726e-03
 -6.26087807e-03  3.02317812e-03  5.73602283e-03 -4.28468460e-04
  6.01001762e-03 -1.69198965e-03  5.43279470e-03  5.62699022e-03
 -2.88454197e-03 -3.30683965e-03  4.35496712e-03  6.48480591e-03
 -5.20145666e-03 -3.04176444e-03 -4.21770568e-03 -3.28701122e-03
  5.91868810e-03  6.96211411e-03  4.04241174e-03 -4.78452628e-03
  6.74957958e-03  6.22083070e-03 -3.88811704e-04  1.82479643e-03
  5.20908605e-03  2.60155604e-03  6.63469193e-03  8.44931789e-04
 -8.29745518e-04  3.60376697e-03  9.97494182e-04  3.64541569e-03
 -6.90072487e-03  3.80146210e-03 -5.36299117e-03 -6.35306719e-03
  1.03268753e-03  3.54132643e-03 -4.01312104e-03  2.90626746e-03
  1.29579833e-03  5.81664084e-03  2.46191290e-03  5.72887267e-03
  5.45967242e-03  4.87193795e-03  6.61741538e-03 -3.24694377e-03
 -3.05439571e-03 -4.30007078e-03 -4.81643057e-03  5.23719897e-03
 -2.92609792e-03  3.11786964e-03  5.06377992e-03  6.93156574e-04
 -6.34295901e-03 -5.20360405e-03  6.53520633e-03 -2.19878290e-03
 -4.23761225e-03 -4.46902040e-03  3.49561972e-03 -3.77943695e-03
 -1.19474523e-03 -5.53529271e-03 -5.21859423e-03  5.27612485e-04
 -3.22711340e-03  1.97883571e-03 -1.37254251e-03 -4.60044753e-03
 -5.11436754e-03 -4.77549801e-03 -3.34637898e-03 -6.94803859e-03
  2.89079417e-03 -7.18483197e-03 -9.93689973e-05 -4.96657772e-03
  6.76707823e-03 -4.62419960e-03  5.78926771e-03 -5.18438170e-03
 -2.05662931e-03  3.27891351e-04 -8.57164219e-04 -4.31960935e-03
 -4.55563453e-03  7.15334687e-03  1.82591525e-03  7.99936920e-05
  2.45142075e-03  5.73190392e-03 -1.54055161e-03 -1.51643201e-03
 -7.22129499e-03 -5.17472970e-04 -8.08219555e-04 -6.38394953e-03
  5.06663481e-03 -5.11019352e-03  3.69304729e-03  1.81648701e-03
 -8.70826108e-04  5.90869986e-03  2.05946655e-03 -4.32638973e-03
 -2.48755354e-03  1.63676471e-03 -7.25196707e-03  4.79333334e-03
  4.19589915e-03  6.51303768e-04  2.76840533e-03 -2.50836736e-03
 -6.46075359e-03  1.25353015e-03  1.33234235e-03 -2.13799161e-04
 -2.87122657e-04 -4.04575035e-03 -5.93910528e-03  1.06039907e-03
  5.28653467e-04 -5.54096450e-03 -6.83464033e-03 -6.89233810e-03
 -3.02152888e-03  1.23447448e-03  2.23110257e-03 -5.61453445e-03
 -5.80583301e-03 -6.01375088e-04  5.93332855e-03  7.31844717e-03
 -4.38679230e-03 -2.48353842e-03 -4.45933690e-03 -4.31983253e-03
  6.04181330e-03 -2.29214770e-03 -2.65686023e-03  1.07060705e-03
  3.17073542e-04  6.97229815e-03 -2.62600050e-03  3.63202727e-03
  3.41508028e-03 -6.49521498e-03 -9.85550166e-04  4.92309670e-03
 -4.84045929e-03 -6.10941181e-03 -6.26433188e-03  1.96075852e-04
  1.70661732e-03  4.63002121e-03 -3.52289210e-03  2.43618431e-03
 -2.64042274e-03 -5.41311502e-03 -7.33775376e-03 -7.14481501e-04
 -6.46917530e-03 -5.40453482e-03  7.22469128e-03  2.06925380e-03
  1.81981272e-03 -9.21539612e-04  5.48254267e-03 -5.33146979e-03
 -3.48952775e-04  3.75635835e-03 -6.01750009e-03  3.11628044e-03
 -6.52624412e-03  3.47105111e-03 -1.11631997e-03 -3.85723350e-03
 -4.40994989e-03  1.17136573e-03 -3.24804979e-03 -2.00483200e-03
  5.87188180e-04  5.16416785e-03  2.25585266e-03  3.65669899e-04
  5.02527363e-04  1.09367410e-03 -2.92367173e-03  4.60333182e-03
  2.98082391e-03  6.09591745e-03 -2.61550086e-04 -3.48777157e-03]
networks.xnet.scale.coeff: torch.Size([1, 9216]) torch.float64
[[0. 0. 0. ... 0. 0. 0.]]
networks.xnet.scale.layer.weight: torch.Size([9216, 256]) torch.float64
[[-0.0314128   0.02457701 -0.01737903 ...  0.05324728  0.02920562
  -0.01480562]
 [ 0.04872326  0.04276607 -0.03822993 ...  0.04910911 -0.00361292
  -0.0327975 ]
 [-0.007368    0.04980131 -0.03215754 ... -0.00088707  0.04168787
   0.01848812]
 ...
 [-0.02830753 -0.05883997  0.00450329 ...  0.05939779  0.03053409
  -0.00791589]
 [-0.03962345 -0.01745788 -0.00161624 ...  0.03181425 -0.04709787
   0.04662044]
 [-0.05086516  0.01152248 -0.02808823 ...  0.05979598  0.0187143
   0.02237458]]
networks.xnet.scale.layer.bias: torch.Size([9216]) torch.float64
[ 0.05768892 -0.05574379 -0.02721741 ...  0.0429844  -0.0245669
  0.01514744]
networks.xnet.transf.coeff: torch.Size([1, 9216]) torch.float64
[[0. 0. 0. ... 0. 0. 0.]]
networks.xnet.transf.layer.weight: torch.Size([9216, 256]) torch.float64
[[-5.48741770e-02  5.88052983e-02 -4.66361299e-02 ...  4.88506894e-03
   4.86617921e-03 -4.32845085e-02]
 [ 1.52577676e-02 -3.61066089e-02 -2.71776527e-02 ... -3.64660092e-02
  -2.36572372e-02  3.22652308e-02]
 [-1.60658234e-02  1.18504295e-02  5.53077906e-02 ...  2.12042919e-03
  -3.51854507e-02 -2.75998519e-02]
 ...
 [ 5.39795249e-02  5.57953445e-02 -5.10586758e-02 ...  1.91429723e-02
   7.08082998e-05 -5.18016647e-02]
 [ 5.80865506e-03 -6.10658241e-02  2.19656541e-02 ...  3.99923136e-02
  -2.62750389e-02 -4.53906247e-02]
 [-4.22715684e-02  2.91019650e-02 -2.99350749e-02 ...  5.93449722e-03
  -2.18851812e-02 -3.72887999e-02]]
networks.xnet.transf.layer.bias: torch.Size([9216]) torch.float64
[ 0.0160626   0.03637902 -0.04792599 ... -0.02090929  0.03092064
 -0.02681802]
networks.xnet.transl.weight: torch.Size([9216, 256]) torch.float64
[[ 0.0597546   0.04724709  0.00974501 ...  0.03109464  0.05561356
  -0.06129477]
 [-0.00993569  0.01303274  0.03429567 ... -0.0537229   0.02269541
   0.0060408 ]
 [-0.03242319  0.05091612 -0.04957046 ...  0.02764103  0.0523916
   0.00046056]
 ...
 [-0.02372713  0.02379183  0.04659715 ... -0.02335574  0.00308037
   0.0329444 ]
 [ 0.04570751 -0.01026316 -0.02490364 ...  0.06235519 -0.0481135
   0.04846663]
 [-0.01367436  0.02093689  0.00793577 ... -0.03375852  0.03164459
  -0.0471489 ]]
networks.xnet.transl.bias: torch.Size([9216]) torch.float64
[-0.05460216 -0.00158335  0.00726456 ... -0.00648457 -0.03945205
  0.03027343]
networks.vnet.input_layer.xlayer.weight: torch.Size([256, 8192]) torch.float64
[[ 9.16479776e-03 -5.02985351e-03 -1.02646308e-02 ... -5.54809175e-03
   2.49656214e-03  5.17212957e-03]
 [-1.10012665e-02  1.43435235e-03  2.52587170e-03 ...  2.85077495e-03
   7.27665005e-03  1.25115602e-04]
 [-4.05602443e-03  5.24941517e-05  2.08143842e-03 ... -8.23178343e-03
  -1.25323771e-03  1.87069762e-03]
 ...
 [ 1.65006943e-03  5.20871048e-03 -1.00319636e-02 ... -8.65131086e-04
  -3.36906845e-03  9.64724941e-03]
 [-4.45201502e-03 -7.09414820e-03  1.06516265e-02 ... -1.70374089e-03
   5.39770888e-03  1.16063326e-03]
 [-2.58663111e-03 -7.89414569e-03 -3.60060986e-03 ...  2.35312618e-03
   3.82747698e-03  6.42668906e-03]]
networks.vnet.input_layer.xlayer.bias: torch.Size([256]) torch.float64
[ 7.19072073e-03  8.54899036e-03 -2.74573756e-04 -1.00589086e-02
 -1.01545971e-02 -4.92664753e-03  3.33148504e-03 -9.15564118e-03
 -2.42099720e-03  2.73065021e-03  9.08042100e-03  5.53534517e-03
 -7.83710822e-03 -7.49573676e-03  6.42186908e-03  8.03173418e-03
 -5.40192935e-03  6.03333723e-03  7.86975860e-03  9.10750021e-03
  7.69085484e-03  5.75742039e-03  5.05968312e-03 -1.08030831e-02
  3.19505866e-03  4.33518487e-03 -8.13443321e-03  3.26326675e-03
 -2.42816999e-04 -4.49394122e-03 -5.71015658e-03  1.04592739e-02
  8.85128136e-03  9.48734877e-03  9.57133407e-04 -4.04794034e-03
 -5.37902552e-04  6.59268074e-03 -8.24567402e-03 -9.85347305e-03
 -3.42285874e-03 -6.53992756e-03  1.22830194e-03  7.85893246e-03
 -5.83300876e-03  9.63272175e-03  1.00944380e-02 -8.23972365e-03
  6.38082055e-03 -8.30738685e-03 -1.28466574e-03 -5.92072715e-03
 -9.94081339e-04 -5.04786137e-03 -4.81081172e-03  7.90118625e-03
  1.71037484e-03  4.65126599e-03  3.37058664e-03 -4.20112172e-03
  3.59920358e-03  7.24500793e-03 -3.96064082e-03  7.38878616e-03
 -4.49233823e-03  1.05992911e-02  6.25592970e-04 -1.42090120e-03
  9.97637162e-03 -5.69488534e-03  8.76500682e-04 -1.01360952e-02
 -2.37918758e-03  9.16485953e-03 -5.39184741e-04 -9.55458902e-03
  2.15706793e-03  6.87956224e-03  9.80659942e-03  1.44044538e-03
 -1.47955633e-03 -6.09390842e-03  3.60479523e-03 -6.82449391e-03
  5.40907839e-03  1.01111670e-03 -6.58945581e-03  2.83824036e-04
  8.00948478e-03  9.06618547e-03  2.53404918e-03 -2.60766025e-03
  8.65071755e-03  4.25651017e-04  5.35529727e-03  1.99133934e-03
  2.17451416e-03  5.14984483e-03 -4.56392209e-03  2.18987254e-03
  2.15989875e-03  2.69606540e-04  3.48712402e-03  1.87305008e-04
 -3.78771696e-03 -1.00614225e-03  3.22000781e-03  1.19988308e-04
  2.42807845e-03  7.89254160e-03  9.15674794e-03  5.91687571e-03
  1.07615647e-02  1.00567930e-03 -7.01280272e-03  2.52361863e-03
 -2.47741325e-03 -1.00704911e-02 -7.94936428e-03  9.14483937e-03
  6.37746665e-03  7.03343740e-03  5.14637055e-03 -5.11663552e-03
 -8.00740179e-03  8.83483394e-03  6.53089076e-03  9.00223946e-03
  3.97538834e-03  6.24483879e-03  7.31143293e-03 -1.00182043e-02
 -9.54410363e-03 -5.58611888e-03 -2.97078564e-03 -4.44242630e-03
 -1.97398918e-03  2.85262394e-03 -2.06799186e-05 -6.70357733e-03
 -2.88790893e-03  2.64525844e-03 -4.77558126e-03  2.39335661e-03
 -2.24044925e-03 -4.06998727e-03  1.06852282e-02 -4.68211211e-03
 -2.39275674e-05  1.78237780e-03  7.35727064e-03  3.83212756e-03
  1.05362754e-02 -1.04227983e-02  5.49080533e-03 -4.21863781e-03
 -8.09427094e-03  4.07019125e-03 -2.45255795e-03 -2.53837092e-03
 -8.00514087e-03  8.73080505e-03 -8.95351911e-04  5.55943184e-03
  1.05709522e-02  1.08791111e-02 -1.27190779e-03 -6.14560145e-03
  3.24457839e-03  5.82486941e-03  8.54307159e-03  1.00030933e-02
 -1.72313442e-03 -8.25437185e-03 -1.09750088e-02  1.56909026e-03
  1.03776179e-02  6.30748559e-03 -6.85598815e-03 -1.12541891e-03
  9.08712741e-03  8.86892888e-03 -6.65631585e-03  4.50625609e-03
  6.52531325e-03 -7.26732740e-03  1.06536875e-03 -4.89295777e-03
  8.56154583e-03 -1.37062236e-03  7.88639999e-03  4.09090142e-03
  2.70268365e-03  2.09226824e-03 -8.10420964e-03 -5.58339073e-03
 -1.20819768e-03  7.11240267e-03  9.43157146e-03 -3.08990524e-03
  8.97867491e-03  3.34746366e-03  1.09295356e-02 -8.33854547e-03
  8.38929734e-04 -3.84791235e-03 -4.82723130e-03 -3.57815038e-03
 -9.57405736e-03 -1.26486527e-03  1.80339331e-04 -9.96310653e-03
 -5.69077756e-03  1.02989043e-02  7.48580783e-03 -7.39925761e-03
  1.63322755e-03  8.09922551e-03  2.04141394e-03 -2.41658959e-03
 -3.70372619e-04  9.03762479e-03 -4.80271241e-03 -7.08458501e-03
 -5.17033153e-03 -5.79945109e-03  1.24089517e-03 -3.33699629e-03
  3.73987379e-03  9.65713624e-03 -4.10453665e-03  4.65230530e-03
  9.41206321e-03  1.98615070e-03  1.89076444e-04 -1.09669487e-02
 -1.05897327e-02 -6.85746166e-03  9.04909639e-03  3.74210626e-03
  7.73146379e-03 -5.16318588e-03 -2.38919870e-03 -8.10861969e-03
 -2.64679669e-03 -2.54794893e-03 -5.72662559e-03  2.21213630e-03
 -3.11922885e-03 -8.05962367e-03 -5.62810358e-03 -4.50267728e-03
  6.65414399e-03  6.62929741e-03  4.13247675e-04  1.00713824e-02]
networks.vnet.input_layer.vlayer.weight: torch.Size([256, 8192]) torch.float64
[[-1.14911505e-04  1.02167470e-03  1.40021033e-03 ...  1.24707499e-03
   7.39021474e-03 -7.66842841e-05]
 [ 1.94857639e-03 -1.01153101e-02  5.93229540e-03 ... -7.92573245e-04
  -4.16384139e-03 -2.18078512e-03]
 [ 6.67728377e-04  2.17963982e-03  5.26611412e-03 ...  9.58798745e-03
   2.09696865e-03 -9.27834153e-03]
 ...
 [ 6.47180548e-03 -1.07005036e-02  3.93737006e-03 ... -5.18451888e-04
  -6.75962666e-03  3.20640579e-03]
 [ 4.78800955e-03  9.73209930e-03 -8.75317563e-03 ... -1.81879817e-03
   4.40209660e-03 -3.12276728e-03]
 [-8.81313864e-03 -3.99257134e-03  8.84710606e-03 ... -6.06736885e-03
  -7.77339244e-04  7.38524802e-03]]
networks.vnet.input_layer.vlayer.bias: torch.Size([256]) torch.float64
[-8.62724045e-03  1.42222980e-03 -6.96917869e-03 -5.58672282e-03
 -1.08436920e-02  1.92805181e-03  9.85077112e-03 -3.70476600e-03
 -1.09324801e-02 -3.16374460e-03  1.03695714e-02  5.06699318e-05
 -4.99672326e-04 -1.99306211e-03 -2.84156206e-03  4.66317161e-03
 -3.03587761e-03  7.51455137e-03 -2.33996069e-03  2.59240772e-05
  7.89245214e-03 -8.62472647e-04  1.73067098e-03  3.98049967e-03
  1.73062241e-03  6.62573647e-03  7.40045855e-03  2.67644125e-03
 -9.08800186e-03  5.89813445e-03  8.16554978e-03  6.63839573e-03
 -9.42590336e-03 -4.87546360e-04 -3.67365451e-03 -7.04534530e-03
  4.69672830e-03  2.60295575e-03 -5.76011135e-03 -6.28032865e-03
  1.91679316e-03  2.98078939e-03  3.62183797e-03 -3.00271988e-03
  1.63915131e-03  2.34721714e-03  6.65671087e-03 -2.31834993e-03
 -9.18060747e-03 -9.18534600e-03 -2.52645532e-03 -5.11630776e-03
 -1.44994545e-03 -1.03017183e-02 -9.78218895e-03 -1.01023624e-02
  1.18473846e-03  1.06675516e-02 -4.46746439e-03 -8.31554290e-03
  3.62939942e-03 -2.85054999e-03  3.50449313e-04  5.88476446e-03
 -7.41086950e-04 -1.88300384e-03 -9.95598461e-03 -3.86232625e-03
  8.51256252e-03 -4.71688463e-03  5.85945792e-03 -4.08366811e-03
  3.74353387e-03 -7.63856755e-03  6.01017018e-03 -2.08397059e-04
 -5.98487160e-03 -2.90322952e-03 -1.70036238e-03 -9.92108061e-03
 -2.22982220e-03  4.83112948e-03 -2.99896346e-03  3.67172178e-03
 -6.98774964e-04 -8.77047812e-03  1.62358117e-03 -1.78590660e-03
  5.38471356e-03 -3.43516886e-03  8.14706400e-03 -5.83946133e-03
 -3.43393746e-03  7.73027246e-03  6.63517574e-04 -1.97393104e-03
 -6.85942244e-03  9.00629517e-03  7.97936788e-03 -9.46204611e-03
 -9.91962740e-04 -1.10322243e-02  2.51705047e-03  9.34333540e-03
 -2.04383936e-03 -3.20343216e-03  9.29710695e-03 -3.91017821e-03
  6.39077050e-03  6.43472117e-03  9.74522830e-03 -6.93840319e-03
 -2.08547302e-03  1.38644474e-03  9.37315176e-03  4.56605920e-03
 -7.89351592e-03  9.38851564e-03  5.63796455e-03 -9.00095278e-03
  5.18531879e-03  5.30712437e-03 -8.30732223e-03 -7.40450098e-03
 -8.20939741e-03 -2.84902854e-03 -4.78296110e-03  4.23491318e-03
 -6.74215327e-04  7.31200359e-03  5.35229365e-03  5.08346832e-03
  4.55305812e-04 -9.43407806e-03 -3.95355860e-03  4.08208203e-03
  9.17557942e-03 -5.82235799e-03 -2.40188538e-03  5.86607196e-03
  9.26253560e-04  3.60195632e-04  5.37108378e-03  9.55347231e-03
  2.21315472e-03  2.83197797e-03  2.42350609e-03  4.04436381e-03
  5.66261568e-03  9.10928763e-03 -9.95752380e-03  9.69473082e-03
 -7.39153944e-04  3.66869476e-03 -1.01362383e-02  9.56476828e-03
 -2.82710541e-04  7.12206993e-03 -5.87929074e-04 -7.01050166e-03
 -4.23527215e-03 -8.87308949e-03 -1.39642809e-04  9.37101367e-03
  9.67923506e-05  1.99839363e-03 -8.40962420e-03  8.90230844e-03
 -1.01463100e-02 -1.92307743e-03  4.56673532e-03 -6.56275075e-03
  9.59734817e-04  1.02384073e-02 -5.02270952e-03  8.87749118e-03
 -1.08131623e-02 -4.46696784e-03  5.71411620e-04 -8.43994833e-03
  4.43985783e-03  1.97354673e-03 -1.01913055e-02 -4.17295006e-03
 -4.20001283e-03 -4.55569830e-03 -4.80213356e-03  4.86272493e-03
  4.45352684e-03 -2.98058407e-04 -2.77590550e-04  7.66098638e-03
  5.67088367e-03 -1.01337152e-02 -8.72584257e-03 -6.31807112e-03
  1.36405857e-03  2.23078281e-03 -7.77014046e-03 -5.62621277e-03
  1.21089554e-03 -8.25528093e-03 -4.31437392e-03  6.83876641e-03
 -2.20724004e-03 -6.96057323e-03  1.10153134e-02  9.20152266e-03
 -5.46451809e-04  3.95414653e-03 -6.12336496e-03 -1.01812970e-02
  2.58442604e-03  3.99579709e-03  2.15366084e-03  3.09138548e-03
 -1.98773575e-03  9.45093202e-03  4.53305893e-03 -4.35642434e-03
 -9.85531045e-04 -2.68568653e-03 -1.70380022e-03 -5.19579726e-03
 -1.04297413e-02  5.08077444e-03  8.74058023e-03  6.00930085e-03
  1.54289208e-03  6.84188667e-03 -9.08237877e-03  4.24600744e-03
  9.33722584e-06 -3.77439462e-03  8.54248635e-03 -7.56578027e-03
 -6.76504009e-03 -5.90847587e-03  4.09332418e-03  3.55476762e-03
  5.98836902e-04 -7.42746846e-03  9.41937684e-03  5.74515692e-03
  9.33460074e-03  1.01823407e-02  4.62239994e-03  4.47014706e-04
 -1.10312440e-02 -1.09556414e-02  1.00085173e-02  7.71189410e-03
 -8.37871661e-03  1.06354341e-03  1.81342397e-03  1.59670919e-03]
networks.vnet.scale.coeff: torch.Size([1, 9216]) torch.float64
[[0. 0. 0. ... 0. 0. 0.]]
networks.vnet.scale.layer.weight: torch.Size([9216, 256]) torch.float64
[[-0.00228215 -0.03766236  0.00310601 ...  0.03574621 -0.03147162
   0.05768579]
 [ 0.04740894  0.05251629  0.05302966 ...  0.03484848 -0.02955205
   0.06163016]
 [ 0.01222568  0.00514609 -0.0334526  ...  0.01297469 -0.02803657
   0.06087524]
 ...
 [ 0.05592883 -0.0102075  -0.05230017 ... -0.00338015 -0.00971749
  -0.02981181]
 [-0.04028498  0.04087614  0.04200986 ... -0.02805969  0.03686787
   0.04125376]
 [ 0.00587553  0.01966032 -0.00264878 ...  0.04938714  0.054406
  -0.00516837]]
networks.vnet.scale.layer.bias: torch.Size([9216]) torch.float64
[ 0.00492612  0.05642332  0.01745884 ...  0.05758128  0.03186792
 -0.00131235]
networks.vnet.transf.coeff: torch.Size([1, 9216]) torch.float64
[[0. 0. 0. ... 0. 0. 0.]]
networks.vnet.transf.layer.weight: torch.Size([9216, 256]) torch.float64
[[-0.04669794 -0.02162502 -0.05433034 ...  0.05776646 -0.04041847
  -0.03207429]
 [ 0.01005234 -0.04020476 -0.04877582 ...  0.05317922  0.01688223
   0.00213256]
 [-0.05261319  0.0362059  -0.04666302 ... -0.02129076 -0.03451092
  -0.01450992]
 ...
 [ 0.00717067  0.04310912  0.04005595 ... -0.01275533 -0.0141606
  -0.04939128]
 [-0.02057675 -0.04490391 -0.0232821  ...  0.02815117 -0.03451967
  -0.03208819]
 [ 0.04748397 -0.05835794 -0.04562179 ...  0.02394428  0.00509771
   0.01316225]]
networks.vnet.transf.layer.bias: torch.Size([9216]) torch.float64
[ 0.00330597  0.02071309  0.00124246 ...  0.03318861 -0.03303616
 -0.04279837]
networks.vnet.transl.weight: torch.Size([9216, 256]) torch.float64
[[ 0.00344245 -0.04238582 -0.04490574 ... -0.02783231 -0.04849146
   0.01404108]
 [ 0.02450462  0.04666956  0.03030368 ...  0.04106765 -0.00071757
   0.00888672]
 [-0.00462819 -0.04935891 -0.008225   ...  0.00929701 -0.01122217
   0.01513675]
 ...
 [ 0.05063408  0.00755534 -0.00132726 ... -0.00643975  0.02790855
   0.01580877]
 [ 0.01441602 -0.00605052  0.05652107 ...  0.01595805 -0.03783348
   0.0410961 ]
 [ 0.00295196 -0.03321374  0.05530547 ...  0.02630581  0.04170922
  -0.04991051]]
networks.vnet.transl.bias: torch.Size([9216]) torch.float64
[ 0.04973598 -0.01250011 -0.00308125 ...  0.04882267  0.00540407
 -0.03680365]
xeps.0: torch.Size([]) torch.float64
0.01
xeps.1: torch.Size([]) torch.float64
0.01
xeps.2: torch.Size([]) torch.float64
0.01
xeps.3: torch.Size([]) torch.float64
0.01
veps.0: torch.Size([]) torch.float64
0.01
veps.1: torch.Size([]) torch.float64
0.01
veps.2: torch.Size([]) torch.float64
0.01
veps.3: torch.Size([]) torch.float64
0.01

[2025-04-30 15:43:03][I][pytorch/trainer:2006:l2hmc.trainers.pytorch.trainer] --------------------------------------------------------------------------------

[2025-04-30 15:43:03][I][ipykernel_24193/3178487732:2:ezpz.log] ExperimentConfig(wandb={'setup': {'id': None, 'group': None, 'config': None, 'save_code': True, 'sync_tensorboard': True, 'mode': 'online', 'resume': 'allow', 'entity': 'l2hmc-qcd', 'project': 'l2hmc-qcd', 'settings': {'start_method': 'thread'}, 'tags': ['beta_init=6.0', 'beta_final=6.0']}}, steps=Steps(nera=1, nepoch=10, test=50, log=1, print=1, extend_last_era=1), framework='pytorch', loss=LossConfig(use_mixed_loss=True, charge_weight=0.0, rmse_weight=0.1, plaq_weight=0.1, aux_weight=0.0), network=NetworkConfig(units=[256], activation_fn='tanh', dropout_prob=0.0, use_batch_norm=False), conv=ConvolutionConfig(filters=[], sizes=[], pool=[]), net_weights=NetWeights(x=NetWeight(s=0.0, t=1.0, q=1.0), v=NetWeight(s=1.0, t=1.0, q=1.0)), dynamics=DynamicsConfig(nchains=8, group='SU3', latvolume=[4, 4, 4, 4], nleapfrog=4, eps=0.01, eps_hmc=0.25, use_ncp=True, verbose=True, eps_fixed=False, use_split_xnets=False, use_separate_networks=False, merge_directions=True), learning_rate=LearningRateConfig(lr_init=0.0001, mode='auto', monitor='loss', patience=5, cooldown=0, warmup=1000, verbose=True, min_lr=1e-06, factor=0.98, min_delta=0.0001, clip_norm=1.0), annealing_schedule=AnnealingSchedule(beta_init=6.0, beta_final=6.0, dynamic=False), gradient_accumulation_steps=1, restore=False, save=False, c1=0.0, port='2345', compile=True, profile=False, init_aim=False, init_wandb=False, use_wandb=False, use_tb=False, debug_mode=False, default_mode=True, print_config=True, precision='float32', ignore_warnings=True, backend='DDP', seed=9992, ds_config_path='/Users/samforeman/projects/saforem2/l2hmc-qcd/src/l2hmc/conf/ds_config.yaml', name=None, width=200, nchains=None, compression=False)

HMC

xhmc, history_hmc = evaluate(
    nsteps=50,
    exp=ptExpSU3,
    beta=6.0,
    x=state.x,
    eps=0.1,
    nleapfrog=8,
    job_type='hmc',
    nlog=1,
    nprint=50,
    grab=True
)

Output:

[2025-04-30 15:43:31][I][pytorch/experiment:117:l2hmc.experiment.pytorch.experiment] Running 50 steps of hmc at beta=6.0000

[2025-04-30 15:43:31][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 0

[2025-04-30 15:43:31][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 1

[2025-04-30 15:43:31][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 2

[2025-04-30 15:43:32][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 3

[2025-04-30 15:43:32][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 4

[2025-04-30 15:43:32][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 5

[2025-04-30 15:43:32][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 6

[2025-04-30 15:43:33][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 7

[2025-04-30 15:43:33][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 8

[2025-04-30 15:43:33][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 9

[2025-04-30 15:43:33][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 10

[2025-04-30 15:43:34][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 11

[2025-04-30 15:43:34][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 12

[2025-04-30 15:43:34][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 13

[2025-04-30 15:43:34][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 14

[2025-04-30 15:43:34][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 15

[2025-04-30 15:43:35][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 16

[2025-04-30 15:43:35][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 17

[2025-04-30 15:43:35][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 18

[2025-04-30 15:43:35][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 19

[2025-04-30 15:43:36][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 20

[2025-04-30 15:43:36][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 21

[2025-04-30 15:43:36][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 22

[2025-04-30 15:43:36][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 23

[2025-04-30 15:43:37][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 24

[2025-04-30 15:43:37][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 25

[2025-04-30 15:43:37][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 26

[2025-04-30 15:43:37][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 27

[2025-04-30 15:43:38][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 28

[2025-04-30 15:43:38][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 29

[2025-04-30 15:43:38][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 30

[2025-04-30 15:43:38][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 31

[2025-04-30 15:43:39][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 32

[2025-04-30 15:43:39][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 33

[2025-04-30 15:43:39][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 34

[2025-04-30 15:43:39][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 35

[2025-04-30 15:43:40][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 36

[2025-04-30 15:43:40][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 37

[2025-04-30 15:43:40][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 38

[2025-04-30 15:43:40][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 39

[2025-04-30 15:43:41][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 40

[2025-04-30 15:43:41][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 41

[2025-04-30 15:43:41][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 42

[2025-04-30 15:43:41][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 43

[2025-04-30 15:43:42][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 44

[2025-04-30 15:43:42][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 45

[2025-04-30 15:43:42][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 46

[2025-04-30 15:43:42][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 47

[2025-04-30 15:43:43][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 48

[2025-04-30 15:43:43][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 49

dataset_hmc = history_hmc.get_dataset()
_ = history_hmc.plot_all(title='HMC')

Output:

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

xhmc = ptExpSU3.trainer.dynamics.unflatten(xhmc)
print(f"checkSU(x_eval): {g.checkSU(xhmc)}")
print(f"checkSU(x_eval): {g.checkSU(g.projectSU(xhmc))}")

Output:

checkSU(x_eval): (tensor[8] f64 x∈[2.314e-16, 5.581e-16] μ=3.985e-16 σ=1.437e-16 [5.220e-16, 5.581e-16, 2.338e-16,
5.152e-16, 5.239e-16, 2.825e-16, 2.314e-16, 3.214e-16], tensor[8] f64 x∈[6.633e-16, 1.660e-15] μ=1.198e-15
σ=3.983e-16 [1.438e-15, 1.657e-15, 6.633e-16, 1.660e-15, 1.449e-15, 8.822e-16, 8.161e-16, 1.014e-15])

checkSU(x_eval): (tensor[8] f64 x∈[2.099e-16, 3.202e-16] μ=2.785e-16 σ=4.123e-17 [3.078e-16, 3.202e-16, 2.099e-16,
3.114e-16, 3.049e-16, 2.593e-16, 2.294e-16, 2.850e-16], tensor[8] f64 x∈[7.639e-16, 9.703e-16] μ=8.714e-16
σ=8.022e-17 [9.119e-16, 9.703e-16, 7.805e-16, 9.469e-16, 9.134e-16, 7.639e-16, 7.940e-16, 8.903e-16])

Training

import time
from l2hmc.utils.history import BaseHistory, summarize_dict

history_train = BaseHistory()
x = state.x
for step in range(100):
# log.info(f'HMC STEP: {step}')
tic = time.perf_counter()
x, metrics_ = ptExpSU3.trainer.train_step(
    (x, state.beta)
)
toc = time.perf_counter()
metrics = {
    'train_step': step,
    'dt': toc - tic,
    **metrics_,
}
if step % 5 == 0:
    avgs = history_train.update(metrics)
    summary = summarize_dict(avgs)
    logger.info(summary)

Output:

[2025-04-30 15:45:44][I][ipykernel_24193/30352159:21:ezpz.log] train_step=0 dt=1.577 energy=33.403 logprob=33.488 logdet=-0.085 sldf=-0.065 sldb=0.086 sld=-0.085 xeps=0.010 veps=0.010 acc=0.131 sumlogdet=0.000 beta=6.000 acc_mask=0.000 loss=88.118 plaqs=-0.004 sinQ=-0.002 intQ=-0.026 dQint=0.000 dQsin=0.000

[2025-04-30 15:45:51][I][ipykernel_24193/30352159:21:ezpz.log] train_step=5 dt=1.256 energy=-372.669 logprob=-372.955 logdet=0.285 sldf=0.191 sldb=-0.217 sld=0.285 xeps=0.010 veps=0.010 acc=1.000 sumlogdet=-0.091 beta=6.000 acc_mask=1.000 loss=-647.866 plaqs=0.038 sinQ=-0.002 intQ=-0.034 dQint=0.009 dQsin=0.001

[2025-04-30 15:45:58][I][ipykernel_24193/30352159:21:ezpz.log] train_step=10 dt=1.308 energy=-780.851 logprob=-781.247 logdet=0.396 sldf=0.254 sldb=-0.276 sld=0.396 xeps=0.011 veps=0.010 acc=0.952 sumlogdet=-0.044 beta=6.000 acc_mask=1.000 loss=-538.541 plaqs=0.088 sinQ=-0.002 intQ=-0.036 dQint=0.017 dQsin=0.001

[2025-04-30 15:46:05][I][ipykernel_24193/30352159:21:ezpz.log] train_step=15 dt=1.487 energy=-1234.642 logprob=-1235.128 logdet=0.486 sldf=0.303 sldb=-0.298 sld=0.486 xeps=0.011 veps=0.010 acc=1.000 sumlogdet=0.011 beta=6.000 acc_mask=1.000 loss=-617.539 plaqs=0.134 sinQ=-0.003 intQ=-0.038 dQint=0.022 dQsin=0.002

[2025-04-30 15:46:13][I][ipykernel_24193/30352159:21:ezpz.log] train_step=20 dt=1.424 energy=-1559.334 logprob=-1559.934 logdet=0.600 sldf=0.370 sldb=-0.363 sld=0.600 xeps=0.012 veps=0.010 acc=1.000 sumlogdet=0.029 beta=6.000 acc_mask=1.000 loss=-450.278 plaqs=0.169 sinQ=0.000 intQ=0.005 dQint=0.016 dQsin=0.001

[2025-04-30 15:46:20][I][ipykernel_24193/30352159:21:ezpz.log] train_step=25 dt=1.611 energy=-1901.485 logprob=-1902.100 logdet=0.615 sldf=0.383 sldb=-0.382 sld=0.615 xeps=0.012 veps=0.010 acc=0.766 sumlogdet=0.082 beta=6.000 acc_mask=0.625 loss=-306.696 plaqs=0.203 sinQ=-0.001 intQ=-0.018 dQint=0.013 dQsin=0.001

[2025-04-30 15:46:28][I][ipykernel_24193/30352159:21:ezpz.log] train_step=30 dt=1.358 energy=-2024.281 logprob=-2025.102 logdet=0.821 sldf=0.522 sldb=-0.530 sld=0.821 xeps=0.013 veps=0.010 acc=0.751 sumlogdet=-0.083 beta=6.000 acc_mask=0.750 loss=-136.896 plaqs=0.224 sinQ=0.001 intQ=0.016 dQint=0.012 dQsin=0.001

[2025-04-30 15:46:36][I][ipykernel_24193/30352159:21:ezpz.log] train_step=35 dt=1.532 energy=-2218.666 logprob=-2219.663 logdet=0.997 sldf=0.624 sldb=-0.624 sld=0.997 xeps=0.013 veps=0.011 acc=0.673 sumlogdet=0.001 beta=6.000 acc_mask=0.625 loss=-26.350 plaqs=0.242 sinQ=0.002 intQ=0.033 dQint=0.012 dQsin=0.001

[2025-04-30 15:46:44][I][ipykernel_24193/30352159:21:ezpz.log] train_step=40 dt=1.553 energy=-2388.089 logprob=-2389.183 logdet=1.093 sldf=0.706 sldb=-0.754 sld=1.093 xeps=0.013 veps=0.011 acc=0.386 sumlogdet=-0.040 beta=6.000 acc_mask=0.375 loss=239.300 plaqs=0.259 sinQ=0.002 intQ=0.029 dQint=0.012 dQsin=0.001

[2025-04-30 15:46:52][I][ipykernel_24193/30352159:21:ezpz.log] train_step=45 dt=1.609 energy=-2497.524 logprob=-2498.931 logdet=1.408 sldf=0.886 sldb=-0.906 sld=1.408 xeps=0.013 veps=0.011 acc=0.414 sumlogdet=0.014 beta=6.000 acc_mask=0.375 loss=76.319 plaqs=0.274 sinQ=0.002 intQ=0.026 dQint=0.003 dQsin=0.000

[2025-04-30 15:46:59][I][ipykernel_24193/30352159:21:ezpz.log] train_step=50 dt=1.623 energy=-2696.907 logprob=-2698.606 logdet=1.699 sldf=1.064 sldb=-1.083 sld=1.699 xeps=0.013 veps=0.012 acc=0.665 sumlogdet=0.006 beta=6.000 acc_mask=0.750 loss=-231.184 plaqs=0.293 sinQ=0.002 intQ=0.036 dQint=0.013 dQsin=0.001

[2025-04-30 15:47:07][I][ipykernel_24193/30352159:21:ezpz.log] train_step=55 dt=1.554 energy=-2865.825 logprob=-2867.770 logdet=1.945 sldf=1.209 sldb=-1.215 sld=1.945 xeps=0.013 veps=0.012 acc=0.770 sumlogdet=0.153 beta=6.000 acc_mask=0.750 loss=-210.427 plaqs=0.311 sinQ=0.003 intQ=0.041 dQint=0.016 dQsin=0.001

[2025-04-30 15:47:15][I][ipykernel_24193/30352159:21:ezpz.log] train_step=60 dt=1.709 energy=-2985.928 logprob=-2987.915 logdet=1.987 sldf=1.251 sldb=-1.296 sld=1.987 xeps=0.013 veps=0.012 acc=1.000 sumlogdet=-0.029 beta=6.000 acc_mask=1.000 loss=-278.412 plaqs=0.323 sinQ=0.003 intQ=0.043 dQint=0.010 dQsin=0.001

[2025-04-30 15:47:23][I][ipykernel_24193/30352159:21:ezpz.log] train_step=65 dt=1.557 energy=-3155.115 logprob=-3157.112 logdet=1.997 sldf=1.252 sldb=-1.281 sld=1.997 xeps=0.013 veps=0.013 acc=1.000 sumlogdet=0.006 beta=6.000 acc_mask=1.000 loss=-363.215 plaqs=0.342 sinQ=0.002 intQ=0.025 dQint=0.021 dQsin=0.001

[2025-04-30 15:47:30][I][ipykernel_24193/30352159:21:ezpz.log] train_step=70 dt=1.474 energy=-3309.328 logprob=-3311.369 logdet=2.041 sldf=1.279 sldb=-1.318 sld=2.041 xeps=0.013 veps=0.013 acc=1.000 sumlogdet=-0.011 beta=6.000 acc_mask=1.000 loss=-374.258 plaqs=0.360 sinQ=-0.000 intQ=-0.003 dQint=0.018 dQsin=0.001

[2025-04-30 15:47:37][I][ipykernel_24193/30352159:21:ezpz.log] train_step=75 dt=1.599 energy=-3493.874 logprob=-3496.069 logdet=2.195 sldf=1.355 sldb=-1.349 sld=2.195 xeps=0.013 veps=0.014 acc=1.000 sumlogdet=0.130 beta=6.000 acc_mask=1.000 loss=-380.764 plaqs=0.378 sinQ=0.001 intQ=0.022 dQint=0.011 dQsin=0.001

[2025-04-30 15:47:45][I][ipykernel_24193/30352159:21:ezpz.log] train_step=80 dt=1.682 energy=-3625.884 logprob=-3627.857 logdet=1.973 sldf=1.234 sldb=-1.251 sld=1.973 xeps=0.013 veps=0.015 acc=1.000 sumlogdet=0.019 beta=6.000 acc_mask=1.000 loss=-442.163 plaqs=0.396 sinQ=-0.000 intQ=-0.003 dQint=0.020 dQsin=0.001

[2025-04-30 15:47:53][I][ipykernel_24193/30352159:21:ezpz.log] train_step=85 dt=1.588 energy=-3841.913 logprob=-3843.908 logdet=1.995 sldf=1.230 sldb=-1.213 sld=1.995 xeps=0.012 veps=0.015 acc=1.000 sumlogdet=0.146 beta=6.000 acc_mask=1.000 loss=-556.948 plaqs=0.416 sinQ=0.000 intQ=0.004 dQint=0.008 dQsin=0.001

[2025-04-30 15:48:01][I][ipykernel_24193/30352159:21:ezpz.log] train_step=90 dt=1.512 energy=-4034.155 logprob=-4035.872 logdet=1.718 sldf=1.085 sldb=-1.128 sld=1.718 xeps=0.012 veps=0.016 acc=1.000 sumlogdet=-0.056 beta=6.000 acc_mask=1.000 loss=-518.971 plaqs=0.435 sinQ=-0.000 intQ=-0.001 dQint=0.016 dQsin=0.001

[2025-04-30 15:48:09][I][ipykernel_24193/30352159:21:ezpz.log] train_step=95 dt=1.604 energy=-4136.090 logprob=-4137.796 logdet=1.707 sldf=1.071 sldb=-1.091 sld=1.707 xeps=0.012 veps=0.016 acc=1.000 sumlogdet=-0.034 beta=6.000 acc_mask=1.000 loss=-613.025 plaqs=0.455 sinQ=-0.001 intQ=-0.008 dQint=0.012 dQsin=0.001

dataset_train = history_train.get_dataset()
_ = history_train.plot_all(
    title='Train',
    num_chains=x.shape[0],
)

Output:

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

Evaluation

# state = ptExpSU3.trainer.dynamics.random_state(6.0)
xeval, history_eval = evaluate(
    nsteps=50,
    exp=ptExpSU3,
    beta=6.0,
# x=state.x,
    job_type='eval',
    nlog=1,
    nprint=50,
    grab=True,
)

Output:

[2025-04-30 15:56:12][I][pytorch/experiment:117:l2hmc.experiment.pytorch.experiment] Running 50 steps of eval at beta=6.0000

[2025-04-30 15:56:12][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 0

[2025-04-30 15:56:12][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 1

[2025-04-30 15:56:13][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 2

[2025-04-30 15:56:14][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 3

[2025-04-30 15:56:14][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 4

[2025-04-30 15:56:15][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 5

[2025-04-30 15:56:15][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 6

[2025-04-30 15:56:16][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 7

[2025-04-30 15:56:16][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 8

[2025-04-30 15:56:17][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 9

[2025-04-30 15:56:17][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 10

[2025-04-30 15:56:18][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 11

[2025-04-30 15:56:18][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 12

[2025-04-30 15:56:19][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 13

[2025-04-30 15:56:19][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 14

[2025-04-30 15:56:20][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 15

[2025-04-30 15:56:20][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 16

[2025-04-30 15:56:21][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 17

[2025-04-30 15:56:21][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 18

[2025-04-30 15:56:22][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 19

[2025-04-30 15:56:22][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 20

[2025-04-30 15:56:23][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 21

[2025-04-30 15:56:23][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 22

[2025-04-30 15:56:24][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 23

[2025-04-30 15:56:24][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 24

[2025-04-30 15:56:24][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 25

[2025-04-30 15:56:25][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 26

[2025-04-30 15:56:25][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 27

[2025-04-30 15:56:26][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 28

[2025-04-30 15:56:26][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 29

[2025-04-30 15:56:27][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 30

[2025-04-30 15:56:27][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 31

[2025-04-30 15:56:28][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 32

[2025-04-30 15:56:28][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 33

[2025-04-30 15:56:29][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 34

[2025-04-30 15:56:29][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 35

[2025-04-30 15:56:29][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 36

[2025-04-30 15:56:30][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 37

[2025-04-30 15:56:30][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 38

[2025-04-30 15:56:31][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 39

[2025-04-30 15:56:32][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 40

[2025-04-30 15:56:32][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 41

[2025-04-30 15:56:33][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 42

[2025-04-30 15:56:33][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 43

[2025-04-30 15:56:34][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 44

[2025-04-30 15:56:34][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 45

[2025-04-30 15:56:35][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 46

[2025-04-30 15:56:35][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 47

[2025-04-30 15:56:36][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 48

[2025-04-30 15:56:36][I][pytorch/experiment:121:l2hmc.experiment.pytorch.experiment] STEP: 49

dataset_eval = history_eval.get_dataset()
_ = history_eval.plot_all(title='Eval')

Output:

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

svg

xeval = ptExpSU3.trainer.dynamics.unflatten(xeval)
logger.info(f"checkSU(x_eval): {g.checkSU(xeval)}")
logger.info(f"checkSU(x_eval): {g.checkSU(g.projectSU(xeval))}")

Output:

[2025-04-30 16:08:39][I][ipykernel_24193/2193937887:2:ezpz.log] checkSU(x_eval): (tensor[8] f64 x∈[1.387e-16, 1.458e-16] μ=1.433e-16 σ=2.311e-18 [1.425e-16, 1.457e-16, 1.387e-16, 1.458e-16, 1.436e-16, 1.422e-16, 1.432e-16, 1.447e-16], tensor[8] f64 x∈[5.173e-16, 7.840e-16] μ=6.551e-16 σ=1.029e-16 [7.840e-16, 6.628e-16, 5.173e-16, 6.288e-16, 6.088e-16, 7.452e-16, 5.299e-16, 7.638e-16])

[2025-04-30 16:08:39][I][ipykernel_24193/2193937887:3:ezpz.log] checkSU(x_eval): (tensor[8] f64 x∈[1.352e-16, 1.451e-16] μ=1.409e-16 σ=3.802e-18 [1.451e-16, 1.352e-16, 1.431e-16, 1.416e-16, 1.366e-16, 1.445e-16, 1.378e-16, 1.430e-16], tensor[8] f64 x∈[4.654e-16, 7.451e-16] μ=5.900e-16 σ=9.667e-17 [7.451e-16, 6.309e-16, 4.961e-16, 6.874e-16, 4.654e-16, 6.100e-16, 5.204e-16, 5.651e-16])

import matplotlib.pyplot as plt
pdiff = dataset_eval.plaqs - dataset_hmc.plaqs
pdiff
import xarray as xr

fig, ax = plt.subplots(figsize=(12, 4))
(pdiff ** 2).plot(ax=ax)  #, robust=True)
ax.set_title(r"$\left|\delta U_{\mu\nu}\right|^{2}$ (HMC - Eval)")
outfile = Path(EVAL_DIR).joinpath('pdiff.svg')
#%xmode fig.savefig(outfile.as_posix(), dpi=400, bbox_inches='tight')
plt.show()

Output:

svg