Load Modules¶
In [1]:
Copied!
# Import modules
import arviz as az
import jax
import matplotlib as plt
import pytensor
import hssm
pytensor.config.floatX = "float32"
jax.config.update("jax_enable_x64", False)
plt.use("Agg")
%matplotlib inline
%config InlineBackend.figure_format='retina'
# hssm.set_floatX("float32")
# Import modules
import arviz as az
import jax
import matplotlib as plt
import pytensor
import hssm
pytensor.config.floatX = "float32"
jax.config.update("jax_enable_x64", False)
plt.use("Agg")
%matplotlib inline
%config InlineBackend.figure_format='retina'
# hssm.set_floatX("float32")
1. Overview of Sequential Sampling Models¶
In [2]:
Copied!
from IPython.display import Image
# Display the image
Image(filename="hssm_tutorial_workshop_1/ssm_overview.png")
from IPython.display import Image
# Display the image
Image(filename="hssm_tutorial_workshop_1/ssm_overview.png")
Out[2]:
In [3]:
Copied!
### Check supported models
hssm.defaults.SupportedModels
### Check supported models
hssm.defaults.SupportedModels
Out[3]:
typing.Literal['ddm', 'ddm_sdv', 'full_ddm', 'angle', 'levy', 'ornstein', 'weibull', 'race_no_bias_angle_4', 'ddm_seq2_no_bias', 'lba3', 'lba2', 'racing_diffusion_3', 'poisson_race', 'softmax_inv_temperature_2', 'softmax_inv_temperature_3']
In [4]:
Copied!
### Details about a particular model
hssm.modelconfig.get_default_model_config("weibull")
##### the bounds within "likelihoods" provide reasonable parameter
##### values (take this with a grain of salt)
### Details about a particular model
hssm.modelconfig.get_default_model_config("weibull")
##### the bounds within "likelihoods" provide reasonable parameter
##### values (take this with a grain of salt)
Out[4]:
{'response': ['rt', 'response'],
'list_params': ['v', 'a', 'z', 't', 'alpha', 'beta'],
'choices': [-1, 1],
'description': None,
'likelihoods': {'approx_differentiable': {'loglik': 'weibull.onnx',
'backend': 'jax',
'default_priors': {},
'bounds': {'v': (-2.5, 2.5),
'a': (0.3, 2.5),
'z': (0.2, 0.8),
't': (0.001, 2.0),
'alpha': (0.31, 4.99),
'beta': (0.31, 6.99)},
'extra_fields': None}}}
In [5]:
Copied!
hssm.modelconfig.get_default_model_config("ddm")
hssm.modelconfig.get_default_model_config("ddm")
Out[5]:
{'response': ['rt', 'response'],
'list_params': ['v', 'a', 'z', 't'],
'choices': [-1, 1],
'description': 'The Drift Diffusion Model (DDM)',
'likelihoods': {'analytical': {'loglik': <function hssm.likelihoods.analytical.logp_ddm(data: numpy.ndarray, v: float, a: float, z: float, t: float, err: float = 1e-15, k_terms: int = 20, epsilon: float = 1e-15) -> numpy.ndarray>,
'backend': None,
'bounds': {'v': (-inf, inf),
'a': (0.0, inf),
'z': (0.0, 1.0),
't': (0.0, inf)},
'default_priors': {'t': {'name': 'HalfNormal', 'sigma': 2.0}},
'extra_fields': None},
'approx_differentiable': {'loglik': 'ddm.onnx',
'backend': 'jax',
'default_priors': {'t': {'name': 'HalfNormal', 'sigma': 2.0}},
'bounds': {'v': (-3.0, 3.0),
'a': (0.3, 2.5),
'z': (0.0, 1.0),
't': (0.0, 2.0)},
'extra_fields': None},
'blackbox': {'loglik': <function hssm.likelihoods.blackbox.hddm_to_hssm.<locals>.outer(data: numpy.ndarray, *args, **kwargs)>,
'backend': None,
'bounds': {'v': (-inf, inf),
'a': (0.0, inf),
'z': (0.0, 1.0),
't': (0.0, inf)},
'default_priors': {'t': {'name': 'HalfNormal', 'sigma': 2.0}},
'extra_fields': None}}}
Notes¶
- there is a difference between "ddm" and "full_ddm"
- the bounds within "likelihoods" provide reasonable param values (but take this with a grain of salt)
- the order of "list_params" is important
2. Regression-based modeling¶
Case Study 1: 1 hierarchical layer (trial-level) with 1 within-subject coefficient)¶
- you have intracranial recording in STN and want to understand the decision-relevant dynamics in the activity
- we assume that activity in STN varies on a trial-by-trial basis in a systematic way
- your Hypothesis: STN is implicated in response cautiousness
- you want to test this hypothesis using a classical DDM
Question to audience: which model parameter does STN modulate?¶
Step 1: Regression-based data simulation (one subjects)¶
In [6]:
Copied!
import numpy as np
import pandas as pd
from ssms.basic_simulators.simulator import simulator
v_true = 0.5
a_true = 1.5
z_true = 0.5
t_true = 0.5
# a changes trial wise
# a_trialwise = np.random.normal(loc=2, scale=0.3, size=1000)
Intercept = 1.5
zSTN = np.random.normal(loc=1, scale=1, size=1000).astype(np.float32)
a_trialwise = Intercept + 0.5 * zSTN
# a changes trial wise
theta_mat = np.zeros((1000, 4))
theta_mat[:, 0] = v_true # v
theta_mat[:, 1] = a_trialwise # a
theta_mat[:, 2] = z_true # z
theta_mat[:, 3] = t_true # t
# simulate data
sim_out_trialwise = simulator(
theta=theta_mat, # parameter_matrix
model="ddm", # specify model (many are included in ssms)
n_samples=1, # number of samples for each set of parameters
# (plays the role of `size` parameter in `hssm.simulate_data`)
)
# Turn into nice dataset
df_ddm_reg_case1 = pd.DataFrame(
np.column_stack(
[sim_out_trialwise["rts"][:, 0], sim_out_trialwise["choices"][:, 0]]
),
columns=["rt", "response"],
)
df_ddm_reg_case1["zSTN"] = zSTN
df_ddm_reg_case1
import numpy as np
import pandas as pd
from ssms.basic_simulators.simulator import simulator
v_true = 0.5
a_true = 1.5
z_true = 0.5
t_true = 0.5
# a changes trial wise
# a_trialwise = np.random.normal(loc=2, scale=0.3, size=1000)
Intercept = 1.5
zSTN = np.random.normal(loc=1, scale=1, size=1000).astype(np.float32)
a_trialwise = Intercept + 0.5 * zSTN
# a changes trial wise
theta_mat = np.zeros((1000, 4))
theta_mat[:, 0] = v_true # v
theta_mat[:, 1] = a_trialwise # a
theta_mat[:, 2] = z_true # z
theta_mat[:, 3] = t_true # t
# simulate data
sim_out_trialwise = simulator(
theta=theta_mat, # parameter_matrix
model="ddm", # specify model (many are included in ssms)
n_samples=1, # number of samples for each set of parameters
# (plays the role of `size` parameter in `hssm.simulate_data`)
)
# Turn into nice dataset
df_ddm_reg_case1 = pd.DataFrame(
np.column_stack(
[sim_out_trialwise["rts"][:, 0], sim_out_trialwise["choices"][:, 0]]
),
columns=["rt", "response"],
)
df_ddm_reg_case1["zSTN"] = zSTN
df_ddm_reg_case1
Out[6]:
| rt | response | zSTN | |
|---|---|---|---|
| 0 | 10.496365 | 1.0 | 2.812863 |
| 1 | 2.163771 | 1.0 | 1.250355 |
| 2 | 2.788312 | 1.0 | 2.861103 |
| 3 | 3.030407 | 1.0 | 2.970084 |
| 4 | 3.987627 | 1.0 | 0.999557 |
| ... | ... | ... | ... |
| 995 | 1.251335 | 1.0 | 0.083161 |
| 996 | 6.996762 | 1.0 | 2.011492 |
| 997 | 6.273996 | 1.0 | 0.308007 |
| 998 | 0.823745 | 1.0 | -1.294699 |
| 999 | 2.683449 | 1.0 | 1.990916 |
1000 rows × 3 columns
Looking at the data, what do you notice?¶
Required Data Structure for HSSM¶
- we need variables labeled as "rt" and "response"
- make sure rt are in seconds
- responses are (1,-1) because you are in a classical DDM setting
- make sure regressors are z-scored (particularly important when you have more than two regressors)
Step 2: Model Setup & Priors¶
In [7]:
Copied!
model_ddm_reg_case1 = hssm.HSSM(
data=df_ddm_reg_case1,
model="ddm",
include=[
{
"name": "a",
"formula": "a ~ 1 + zSTN",
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
},
"link": "identity",
}
],
)
model_ddm_reg_case1 = hssm.HSSM(
data=df_ddm_reg_case1,
model="ddm",
include=[
{
"name": "a",
"formula": "a ~ 1 + zSTN",
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
},
"link": "identity",
}
],
)
Model initialized successfully.
Notes to Step 2¶
- make sure your priors are reasonable (always double-check)
- see my blog for useful notes for defining priors on each of these params https://www.gingjehli.com/single-post/choosing-effective-samplers-and-setting-priors
- priors are particularly important in these more complex regression-based models
- it can also be helpful to set "init_val" for the sampler so it starts in a reasonable region.
Step 3: Double-check model specification¶
In [8]:
Copied!
# graphical illustration of model
model_ddm_reg_case1.graph()
# graphical illustration of model
model_ddm_reg_case1.graph()
Out[8]:
Step 4: Sampling from the Posterior Distribution¶
In [9]:
Copied!
# suppress warnings
import warnings
warnings.filterwarnings("ignore")
samples_model_ddm_reg_case1 = model_ddm_reg_case1.sample(
sampler="numpyro",
cores=2,
chains=2,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
# suppress warnings
import warnings
warnings.filterwarnings("ignore")
samples_model_ddm_reg_case1 = model_ddm_reg_case1.sample(
sampler="numpyro",
cores=2,
chains=2,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
Using default initvals.
NUTS[numpyro]: [v, z, t, a_Intercept, a_zSTN]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:00<00:13, 68.96it/s, 31 steps of size 1.43e-02. acc. prob=0.74]
warmup: 10%|█ | 100/1000 [00:00<00:06, 132.15it/s, 7 steps of size 2.22e-02. acc. prob=0.77]
warmup: 15%|█▌ | 150/1000 [00:01<00:09, 86.22it/s, 123 steps of size 3.15e-02. acc. prob=0.77]
warmup: 20%|██ | 200/1000 [00:01<00:07, 113.44it/s, 15 steps of size 2.92e-01. acc. prob=0.78]
warmup: 25%|██▌ | 250/1000 [00:02<00:05, 145.69it/s, 7 steps of size 7.89e-02. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:02<00:03, 180.82it/s, 15 steps of size 3.92e-01. acc. prob=0.78]
warmup: 35%|███▌ | 350/1000 [00:02<00:02, 226.48it/s, 3 steps of size 8.19e-01. acc. prob=0.79]
warmup: 40%|████ | 400/1000 [00:02<00:02, 260.14it/s, 15 steps of size 4.12e-01. acc. prob=0.79]
warmup: 45%|████▌ | 451/1000 [00:02<00:01, 306.60it/s, 7 steps of size 4.77e+00. acc. prob=0.79]
sample: 50%|█████ | 501/1000 [00:02<00:01, 329.87it/s, 15 steps of size 3.63e-01. acc. prob=0.97]
sample: 55%|█████▌ | 551/1000 [00:02<00:01, 366.88it/s, 15 steps of size 3.63e-01. acc. prob=0.85]
sample: 60%|██████ | 601/1000 [00:02<00:01, 376.87it/s, 7 steps of size 3.63e-01. acc. prob=0.87]
sample: 65%|██████▌ | 651/1000 [00:03<00:00, 379.27it/s, 7 steps of size 3.63e-01. acc. prob=0.89]
sample: 70%|███████ | 702/1000 [00:03<00:00, 408.36it/s, 15 steps of size 3.63e-01. acc. prob=0.90]
sample: 75%|███████▌ | 752/1000 [00:03<00:00, 423.81it/s, 7 steps of size 3.63e-01. acc. prob=0.90]
sample: 80%|████████ | 802/1000 [00:03<00:00, 420.67it/s, 15 steps of size 3.63e-01. acc. prob=0.90]
sample: 85%|████████▌ | 852/1000 [00:03<00:00, 423.73it/s, 15 steps of size 3.63e-01. acc. prob=0.90]
sample: 90%|█████████ | 902/1000 [00:03<00:00, 427.78it/s, 15 steps of size 3.63e-01. acc. prob=0.90]
sample: 95%|█████████▌| 952/1000 [00:03<00:00, 423.26it/s, 15 steps of size 3.63e-01. acc. prob=0.90]
sample: 100%|██████████| 1000/1000 [00:03<00:00, 260.95it/s, 3 steps of size 3.63e-01. acc. prob=0.90]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 29%|██▉ | 288/1000 [00:00<00:00, 2876.55it/s, 1 steps of size 1.18e-38. acc. prob=0.00]
sample: 58%|█████▊ | 583/1000 [00:00<00:00, 2916.77it/s, 1 steps of size 1.18e-38. acc. prob=0.00]
sample: 88%|████████▊ | 875/1000 [00:00<00:00, 2916.83it/s, 1 steps of size 1.18e-38. acc. prob=0.00]
sample: 100%|██████████| 1000/1000 [00:00<00:00, 2912.21it/s, 1 steps of size 1.18e-38. acc. prob=0.00]
There were 504 divergences after tuning. Increase `target_accept` or reparameterize.
We recommend running at least 4 chains for robust computation of convergence diagnostics
The rhat statistic is larger than 1.01 for some parameters. This indicates problems during sampling. See https://arxiv.org/abs/1903.08008 for details
The effective sample size per chain is smaller than 100 for some parameters. A higher number is needed for reliable rhat and ess computation. See https://arxiv.org/abs/1903.08008 for details
Notes to Step 4¶
- in a real analysis, you would want to run more chains (usually >=4)
- if models don't converge, as a first step, it is typically a good idea to increase the chain length
- depending on task structure, some model params are more/less likely to trade-off (e.g., simple task design,
t&aare often anti-correlated) - if there are parameter trade-offs or depending on your regression equation, it can be helpful to play around with the sampler
- though: numpyro sampler does not work with blackbox likelihoods.
- sometime variational inference (VI) can be a helpful alternative to MCMC (check out the dedicated tutorials in the HSSM documentation)
Step 5: Model Validation¶
What do you think about these results below (also given Jensen's part)?¶
In [10]:
Copied!
#### quick check on posterior statistics
az.summary(model_ddm_reg_case1.traces)
#### quick check on posterior statistics
az.summary(model_ddm_reg_case1.traces)
Out[10]:
| mean | sd | eti89_lb | eti89_ub | ess_bulk | ess_tail | r_hat | mcse_mean | mcse_sd | |
|---|---|---|---|---|---|---|---|---|---|
| a_zSTN | -0.1 | 0.64 | -0.73 | 0.58 | 2 | 2 | 2.25 | 0.45 | 0.011 |
| v | 0.5 | 0.08 | 0.44 | 0.63 | 2 | 2 | 2.24 | 0.054 | 0.014 |
| z | 0.6 | 0.1 | 0.47 | 0.69 | 2 | 22 | 2.24 | 0.07 | 0.0072 |
| t | 0.3 | 0.24 | 0.014 | 0.51 | 2 | 2 | 2.23 | 0.17 | 0.0096 |
| a_Intercept | 1.8 | 0.24 | 1.5 | 2 | 2 | 23 | 2.24 | 0.16 | 0.016 |
In [11]:
Copied!
#### trace plots
az.plot_trace_dist(model_ddm_reg_case1.traces)
#### trace plots
az.plot_trace_dist(model_ddm_reg_case1.traces)
Out[11]:
<arviz_plots.plot_collection.PlotCollection at 0x1277a4830>
Case Study 2: Participant hierarchy on two within-subject coefficients¶
- besides from STN, we also have an indirect nogo pathway via GPe
- this time, we want to assume that activity in these components modulates drift rate
- even though there is evidence that STN modulates the boundary, we always want to check different model specifications
Step 1: Regression-based data simulation (multiple subjects)¶
In [12]:
Copied!
# Function to simulate data for one participant
def simulate_participant(participant_id, size=300):
"""Simulate DDM trial data for one participant."""
intercept = 1.5
zSTN = np.random.normal(loc=1, scale=2, size=size)
zGPe = np.random.normal(loc=1, scale=2, size=size)
v = intercept + 0.8 * zSTN + 0.3 * zGPe
# Assume `hssm.simulate_data` returns a DataFrame
true_values = np.column_stack(
[v, np.repeat([[1.5, 0.5, 0.5]], axis=0, repeats=300)]
)
dataset_reg_v = hssm.simulate_data(
model="ddm",
theta=true_values,
size=1, # Generate one data point for each of the 1000 set of true values
)
# Adding additional variables to the dataset
dataset_reg_v["zSTN"] = zSTN
dataset_reg_v["zGPe"] = zGPe
dataset_reg_v["participant_id"] = str(participant_id)
return dataset_reg_v
# Simulate data for 10 participants
# Combine datasets into one DataFrame
combined_dataset = pd.concat(
[simulate_participant(i, size=300) for i in range(1, 11)], ignore_index=True
)
combined_dataset
# Function to simulate data for one participant
def simulate_participant(participant_id, size=300):
"""Simulate DDM trial data for one participant."""
intercept = 1.5
zSTN = np.random.normal(loc=1, scale=2, size=size)
zGPe = np.random.normal(loc=1, scale=2, size=size)
v = intercept + 0.8 * zSTN + 0.3 * zGPe
# Assume `hssm.simulate_data` returns a DataFrame
true_values = np.column_stack(
[v, np.repeat([[1.5, 0.5, 0.5]], axis=0, repeats=300)]
)
dataset_reg_v = hssm.simulate_data(
model="ddm",
theta=true_values,
size=1, # Generate one data point for each of the 1000 set of true values
)
# Adding additional variables to the dataset
dataset_reg_v["zSTN"] = zSTN
dataset_reg_v["zGPe"] = zGPe
dataset_reg_v["participant_id"] = str(participant_id)
return dataset_reg_v
# Simulate data for 10 participants
# Combine datasets into one DataFrame
combined_dataset = pd.concat(
[simulate_participant(i, size=300) for i in range(1, 11)], ignore_index=True
)
combined_dataset
Out[12]:
| rt | response | zSTN | zGPe | participant_id | |
|---|---|---|---|---|---|
| 0 | 0.747980 | 1.0 | -0.158168 | 1.413627 | 1 |
| 1 | 1.233290 | 1.0 | 0.365957 | 2.550413 | 1 |
| 2 | 1.078334 | 1.0 | 2.792268 | 1.212261 | 1 |
| 3 | 0.988373 | 1.0 | 2.902137 | -1.633718 | 1 |
| 4 | 1.382792 | 1.0 | 2.014373 | -2.116652 | 1 |
| ... | ... | ... | ... | ... | ... |
| 2995 | 0.784602 | 1.0 | 2.236697 | 2.210372 | 10 |
| 2996 | 0.775152 | 1.0 | 3.069456 | 6.258231 | 10 |
| 2997 | 0.725300 | 1.0 | 4.501823 | 0.557991 | 10 |
| 2998 | 2.103001 | 1.0 | 0.337555 | -2.215638 | 10 |
| 2999 | 0.904685 | 1.0 | 1.697219 | 0.776832 | 10 |
3000 rows × 5 columns
Step 2: Model Setup & Priors¶
Alternative 1: A centered model¶
- specification where the subject-specific coefficients are coming from one hierarchical prior (mother distribution with mu and sigma)
- we can also call it as "0+" models
- this is the specification used in HDDM
- in this case, it's important to set "noncentered=False" (by default it's set to "True")
- The group distributions here will be recovered as v_1|subject_mu and v_x|subject_mu
In [13]:
Copied!
model_reg_v_ex2_A1 = hssm.HSSM(
data=combined_dataset,
model="ddm",
include=[
{
"name": "v",
"formula": "v ~ 0 + (1 + zSTN + zGPe | participant_id)",
"prior": {
"1|participant_id": {
"name": "Normal",
"mu": {"name": "Normal", "mu": 1, "sigma": 1.0, "initval": 0},
"sigma": {"name": "HalfNormal", "sigma": 1, "initval": 0.25},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": {"name": "Normal", "mu": 0, "sigma": 1.0, "initval": 0},
"sigma": {"name": "HalfNormal", "sigma": 1, "initval": 0.25},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": {"name": "Normal", "mu": 0, "sigma": 1.0, "initval": 0},
"sigma": {"name": "HalfNormal", "sigma": 1, "initval": 0.25},
},
},
"link": "identity",
}
],
noncentered=False,
p_outlier=0.05,
)
model_reg_v_ex2_A1 = hssm.HSSM(
data=combined_dataset,
model="ddm",
include=[
{
"name": "v",
"formula": "v ~ 0 + (1 + zSTN + zGPe | participant_id)",
"prior": {
"1|participant_id": {
"name": "Normal",
"mu": {"name": "Normal", "mu": 1, "sigma": 1.0, "initval": 0},
"sigma": {"name": "HalfNormal", "sigma": 1, "initval": 0.25},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": {"name": "Normal", "mu": 0, "sigma": 1.0, "initval": 0},
"sigma": {"name": "HalfNormal", "sigma": 1, "initval": 0.25},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": {"name": "Normal", "mu": 0, "sigma": 1.0, "initval": 0},
"sigma": {"name": "HalfNormal", "sigma": 1, "initval": 0.25},
},
},
"link": "identity",
}
],
noncentered=False,
p_outlier=0.05,
)
Model initialized successfully.
In [14]:
Copied!
# graphical illustration of model
model_reg_v_ex2_A1.graph()
# graphical illustration of model
model_reg_v_ex2_A1.graph()
Out[14]:
Alternative 2: A non-centered model¶
- Non-centered means that each subject is estimated as offset from the group mean
- in this case the group means will be zSTN and zGPe.
- The priors are different here because now we want the mu for the random subject effects to be 0
- the mu should not be estimated from a hyperprior simultaneously with the fixed effects, as those will tradeoff and could produce convergence issues - we still estimate the sigmas with a hyperprior to allow us to get indiv params
- now we keep noncentered=true.
In [15]:
Copied!
model_reg_v_ex2_A2 = hssm.HSSM(
data=combined_dataset,
include=[
{
"name": "v",
"formula": "v ~ 1 + zSTN + zGPe + (1 + zSTN + zGPe | participant_id)",
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.0, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
model_reg_v_ex2_A2 = hssm.HSSM(
data=combined_dataset,
include=[
{
"name": "v",
"formula": "v ~ 1 + zSTN + zGPe + (1 + zSTN + zGPe | participant_id)",
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.0, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
Model initialized successfully.
In [16]:
Copied!
model_reg_v_ex2_A2.graph()
model_reg_v_ex2_A2.graph()
Out[16]:
Important note¶
Note that in this non-centered model, you will need to specify each individual coefficient's prior (at least for your regression equation). See example below
Step 4: Sampling from the Posterior Distribution (Bayesian Model Fitting based on MCMC Procedure)¶
In [17]:
Copied!
## centered model version
samples_model_reg_v_ex2_A1 = model_reg_v_ex2_A1.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
## centered model version
samples_model_reg_v_ex2_A1 = model_reg_v_ex2_A1.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
Using default initvals.
NUTS[numpyro]: [a, z, t, v_1|participant_id_mu, v_1|participant_id_sigma, v_1|participant_id, v_zSTN|participant_id_mu, v_zSTN|participant_id_sigma, v_zSTN|participant_id, v_zGPe|participant_id_mu, v_zGPe|participant_id_sigma, v_zGPe|participant_id]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:02<00:40, 23.70it/s, 31 steps of size 5.94e-03. acc. prob=0.74]
warmup: 10%|█ | 100/1000 [00:05<00:50, 17.83it/s, 255 steps of size 3.30e-03. acc. prob=0.76]
warmup: 15%|█▌ | 150/1000 [00:07<00:40, 20.97it/s, 15 steps of size 4.58e-01. acc. prob=0.78]
warmup: 20%|██ | 200/1000 [00:08<00:30, 25.82it/s, 63 steps of size 1.04e-01. acc. prob=0.78]
warmup: 25%|██▌ | 250/1000 [00:09<00:21, 34.61it/s, 15 steps of size 2.28e-01. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:10<00:19, 36.02it/s, 63 steps of size 7.64e-02. acc. prob=0.78]
warmup: 35%|███▌ | 350/1000 [00:11<00:17, 36.78it/s, 15 steps of size 3.25e-01. acc. prob=0.78]
warmup: 40%|████ | 400/1000 [00:12<00:14, 41.79it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 45%|████▌ | 450/1000 [00:13<00:12, 45.19it/s, 63 steps of size 8.23e-02. acc. prob=0.78]
warmup: 50%|█████ | 500/1000 [00:14<00:10, 46.89it/s, 31 steps of size 1.83e-01. acc. prob=0.78]
sample: 55%|█████▌ | 550/1000 [00:15<00:08, 50.90it/s, 15 steps of size 1.83e-01. acc. prob=0.89]
sample: 60%|██████ | 600/1000 [00:16<00:07, 54.12it/s, 31 steps of size 1.83e-01. acc. prob=0.87]
sample: 65%|██████▌ | 650/1000 [00:16<00:06, 56.82it/s, 15 steps of size 1.83e-01. acc. prob=0.87]
sample: 70%|███████ | 700/1000 [00:17<00:05, 59.56it/s, 31 steps of size 1.83e-01. acc. prob=0.85]
sample: 75%|███████▌ | 750/1000 [00:18<00:04, 60.19it/s, 31 steps of size 1.83e-01. acc. prob=0.85]
sample: 80%|████████ | 800/1000 [00:19<00:03, 56.83it/s, 31 steps of size 1.83e-01. acc. prob=0.86]
sample: 85%|████████▌ | 850/1000 [00:20<00:02, 58.04it/s, 9 steps of size 1.83e-01. acc. prob=0.87]
sample: 90%|█████████ | 900/1000 [00:20<00:01, 58.61it/s, 31 steps of size 1.83e-01. acc. prob=0.85]
sample: 95%|█████████▌| 950/1000 [00:21<00:00, 62.49it/s, 15 steps of size 1.83e-01. acc. prob=0.83]
sample: 100%|██████████| 1000/1000 [00:22<00:00, 63.21it/s, 31 steps of size 1.83e-01. acc. prob=0.83]
sample: 100%|██████████| 1000/1000 [00:22<00:00, 44.58it/s, 31 steps of size 1.83e-01. acc. prob=0.83]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:02<00:39, 24.16it/s, 255 steps of size 3.88e-03. acc. prob=0.73]
warmup: 10%|█ | 100/1000 [00:05<00:46, 19.40it/s, 63 steps of size 3.82e-03. acc. prob=0.76]
warmup: 15%|█▌ | 150/1000 [00:07<00:41, 20.24it/s, 3 steps of size 7.14e-02. acc. prob=0.77]
warmup: 20%|██ | 200/1000 [00:08<00:29, 26.69it/s, 63 steps of size 1.69e-01. acc. prob=0.78]
warmup: 25%|██▌ | 250/1000 [00:09<00:22, 33.40it/s, 15 steps of size 1.82e-01. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:10<00:20, 34.95it/s, 15 steps of size 2.29e-01. acc. prob=0.78]
warmup: 35%|███▌ | 350/1000 [00:11<00:17, 38.09it/s, 15 steps of size 3.68e-01. acc. prob=0.78]
warmup: 40%|████ | 400/1000 [00:12<00:13, 45.14it/s, 31 steps of size 2.11e-01. acc. prob=0.78]
warmup: 45%|████▌ | 450/1000 [00:12<00:10, 50.76it/s, 63 steps of size 7.85e-02. acc. prob=0.78]
warmup: 50%|█████ | 500/1000 [00:13<00:10, 49.01it/s, 15 steps of size 2.03e-01. acc. prob=0.79]
sample: 55%|█████▌ | 550/1000 [00:14<00:07, 56.65it/s, 15 steps of size 2.03e-01. acc. prob=0.87]
sample: 60%|██████ | 600/1000 [00:15<00:06, 62.63it/s, 15 steps of size 2.03e-01. acc. prob=0.82]
sample: 65%|██████▌ | 650/1000 [00:15<00:05, 66.61it/s, 15 steps of size 2.03e-01. acc. prob=0.81]
sample: 70%|███████ | 700/1000 [00:16<00:04, 68.06it/s, 15 steps of size 2.03e-01. acc. prob=0.82]
sample: 75%|███████▌ | 750/1000 [00:17<00:03, 69.21it/s, 15 steps of size 2.03e-01. acc. prob=0.82]
sample: 80%|████████ | 800/1000 [00:17<00:02, 68.29it/s, 31 steps of size 2.03e-01. acc. prob=0.82]
sample: 85%|████████▌ | 850/1000 [00:18<00:02, 73.38it/s, 31 steps of size 2.03e-01. acc. prob=0.78]
sample: 90%|█████████ | 900/1000 [00:19<00:01, 75.59it/s, 15 steps of size 2.03e-01. acc. prob=0.79]
sample: 95%|█████████▌| 950/1000 [00:19<00:00, 77.78it/s, 15 steps of size 2.03e-01. acc. prob=0.78]
sample: 100%|██████████| 1000/1000 [00:20<00:00, 77.04it/s, 15 steps of size 2.03e-01. acc. prob=0.75]
sample: 100%|██████████| 1000/1000 [00:20<00:00, 49.04it/s, 15 steps of size 2.03e-01. acc. prob=0.75]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:02<00:55, 17.12it/s, 383 steps of size 2.19e-03. acc. prob=0.73]
warmup: 10%|█ | 100/1000 [00:06<00:59, 15.08it/s, 5 steps of size 2.58e-03. acc. prob=0.75]
warmup: 15%|█▌ | 150/1000 [00:08<00:43, 19.65it/s, 63 steps of size 7.92e-02. acc. prob=0.77]
warmup: 20%|██ | 200/1000 [00:09<00:31, 25.47it/s, 15 steps of size 1.55e-01. acc. prob=0.78]
warmup: 25%|██▌ | 250/1000 [00:10<00:23, 32.09it/s, 31 steps of size 2.32e-01. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:10<00:18, 37.07it/s, 31 steps of size 2.61e-01. acc. prob=0.78]
warmup: 35%|███▌ | 350/1000 [00:11<00:15, 41.08it/s, 15 steps of size 1.38e-01. acc. prob=0.78]
warmup: 40%|████ | 400/1000 [00:12<00:12, 47.04it/s, 127 steps of size 7.32e-02. acc. prob=0.78]
warmup: 45%|████▌ | 450/1000 [00:13<00:11, 48.86it/s, 15 steps of size 3.71e-01. acc. prob=0.79]
warmup: 50%|█████ | 500/1000 [00:14<00:10, 49.16it/s, 15 steps of size 1.67e-01. acc. prob=0.79]
sample: 55%|█████▌ | 550/1000 [00:15<00:08, 50.16it/s, 50 steps of size 1.67e-01. acc. prob=0.82]
sample: 60%|██████ | 600/1000 [00:16<00:07, 53.28it/s, 15 steps of size 1.67e-01. acc. prob=0.78]
sample: 65%|██████▌ | 650/1000 [00:17<00:06, 55.62it/s, 31 steps of size 1.67e-01. acc. prob=0.77]
sample: 70%|███████ | 700/1000 [00:18<00:05, 56.46it/s, 15 steps of size 1.67e-01. acc. prob=0.79]
sample: 75%|███████▌ | 750/1000 [00:18<00:04, 58.44it/s, 15 steps of size 1.67e-01. acc. prob=0.81]
sample: 80%|████████ | 800/1000 [00:19<00:03, 59.24it/s, 15 steps of size 1.67e-01. acc. prob=0.81]
sample: 85%|████████▌ | 850/1000 [00:20<00:02, 55.45it/s, 15 steps of size 1.67e-01. acc. prob=0.82]
sample: 90%|█████████ | 900/1000 [00:21<00:01, 55.78it/s, 31 steps of size 1.67e-01. acc. prob=0.83]
sample: 95%|█████████▌| 950/1000 [00:22<00:00, 60.30it/s, 31 steps of size 1.67e-01. acc. prob=0.81]
sample: 100%|██████████| 1000/1000 [00:22<00:00, 63.43it/s, 63 steps of size 1.67e-01. acc. prob=0.81]
sample: 100%|██████████| 1000/1000 [00:22<00:00, 43.63it/s, 63 steps of size 1.67e-01. acc. prob=0.81]
There were 57 divergences after tuning. Increase `target_accept` or reparameterize.
We recommend running at least 4 chains for robust computation of convergence diagnostics
The rhat statistic is larger than 1.01 for some parameters. This indicates problems during sampling. See https://arxiv.org/abs/1903.08008 for details
The effective sample size per chain is smaller than 100 for some parameters. A higher number is needed for reliable rhat and ess computation. See https://arxiv.org/abs/1903.08008 for details
In [18]:
Copied!
## centered model version
samples_model_reg_v_ex2_A2 = model_reg_v_ex2_A2.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
## centered model version
samples_model_reg_v_ex2_A2 = model_reg_v_ex2_A2.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
Using default initvals.
NUTS[numpyro]: [a, z, t, v_Intercept, v_zSTN, v_zGPe, v_1|participant_id_sigma, v_1|participant_id_offset, v_zSTN|participant_id_sigma, v_zSTN|participant_id_offset, v_zGPe|participant_id_sigma, v_zGPe|participant_id_offset]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:15<04:47, 3.31it/s, 1023 steps of size 2.55e-03. acc. prob=0.73]
warmup: 5%|▌ | 50/1000 [00:29<04:47, 3.31it/s, 511 steps of size 4.66e-03. acc. prob=0.75]
warmup: 9%|▊ | 86/1000 [00:29<05:25, 2.81it/s, 1023 steps of size 4.25e-03. acc. prob=0.75]
warmup: 9%|▊ | 87/1000 [00:30<05:29, 2.77it/s, 1023 steps of size 6.43e-03. acc. prob=0.76]
warmup: 9%|▉ | 88/1000 [00:30<05:27, 2.78it/s, 511 steps of size 4.23e-03. acc. prob=0.75]
warmup: 9%|▉ | 89/1000 [00:31<05:35, 2.72it/s, 1023 steps of size 4.82e-03. acc. prob=0.75]
warmup: 9%|▉ | 90/1000 [00:31<05:31, 2.75it/s, 511 steps of size 4.89e-03. acc. prob=0.76]
warmup: 9%|▉ | 91/1000 [00:31<05:26, 2.78it/s, 511 steps of size 6.54e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:32<05:21, 2.82it/s, 1023 steps of size 2.36e-03. acc. prob=0.75]
warmup: 9%|▉ | 94/1000 [00:33<05:43, 2.64it/s, 1023 steps of size 3.54e-03. acc. prob=0.75]
warmup: 10%|▉ | 95/1000 [00:33<06:07, 2.46it/s, 1023 steps of size 4.80e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:34<06:32, 2.30it/s, 1023 steps of size 5.10e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:34<06:11, 2.43it/s, 511 steps of size 6.77e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:35<05:54, 2.54it/s, 1023 steps of size 5.33e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:35<05:36, 2.67it/s, 511 steps of size 5.86e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:36<06:18, 2.38it/s, 1023 steps of size 8.43e-02. acc. prob=0.76]
warmup: 11%|█ | 106/1000 [00:36<02:42, 5.49it/s, 127 steps of size 1.06e-01. acc. prob=0.76]
warmup: 11%|█ | 112/1000 [00:36<01:27, 10.21it/s, 31 steps of size 2.06e-01. acc. prob=0.77]
warmup: 12%|█▏ | 120/1000 [00:36<00:52, 16.78it/s, 127 steps of size 9.18e-02. acc. prob=0.77]
warmup: 12%|█▎ | 125/1000 [00:36<00:42, 20.73it/s, 63 steps of size 1.11e-01. acc. prob=0.77]
warmup: 13%|█▎ | 130/1000 [00:36<00:35, 24.52it/s, 63 steps of size 9.43e-02. acc. prob=0.77]
warmup: 14%|█▎ | 135/1000 [00:36<00:30, 28.40it/s, 31 steps of size 2.82e-01. acc. prob=0.77]
warmup: 14%|█▍ | 140/1000 [00:37<00:28, 30.55it/s, 63 steps of size 1.53e-01. acc. prob=0.77]
warmup: 15%|█▍ | 146/1000 [00:37<00:23, 36.46it/s, 63 steps of size 1.39e-01. acc. prob=0.77]
warmup: 15%|█▌ | 151/1000 [00:37<00:22, 37.83it/s, 63 steps of size 1.21e+00. acc. prob=0.78]
warmup: 16%|█▌ | 161/1000 [00:37<00:17, 48.72it/s, 63 steps of size 9.50e-02. acc. prob=0.77]
warmup: 17%|█▋ | 169/1000 [00:37<00:15, 54.55it/s, 63 steps of size 7.44e-02. acc. prob=0.77]
warmup: 17%|█▋ | 174/1000 [00:37<00:15, 52.73it/s, 63 steps of size 8.76e-02. acc. prob=0.77]
warmup: 18%|█▊ | 179/1000 [00:37<00:15, 51.73it/s, 15 steps of size 2.63e-01. acc. prob=0.78]
warmup: 18%|█▊ | 185/1000 [00:37<00:15, 53.72it/s, 15 steps of size 2.82e-01. acc. prob=0.78]
warmup: 19%|█▉ | 192/1000 [00:38<00:13, 57.80it/s, 15 steps of size 3.55e-01. acc. prob=0.78]
warmup: 20%|█▉ | 196/1000 [00:38<00:15, 52.75it/s, 31 steps of size 2.02e-01. acc. prob=0.78]
warmup: 21%|██ | 206/1000 [00:38<00:12, 65.56it/s, 15 steps of size 1.89e-01. acc. prob=0.78]
warmup: 21%|██ | 212/1000 [00:38<00:12, 62.58it/s, 63 steps of size 1.21e-01. acc. prob=0.78]
warmup: 22%|██▏ | 217/1000 [00:38<00:13, 58.63it/s, 63 steps of size 1.16e-01. acc. prob=0.78]
warmup: 22%|██▏ | 224/1000 [00:38<00:12, 60.60it/s, 31 steps of size 2.34e-01. acc. prob=0.78]
warmup: 24%|██▎ | 236/1000 [00:38<00:10, 76.16it/s, 15 steps of size 1.31e-01. acc. prob=0.78]
warmup: 24%|██▍ | 242/1000 [00:38<00:10, 69.46it/s, 63 steps of size 1.41e-01. acc. prob=0.78]
warmup: 25%|██▍ | 249/1000 [00:38<00:10, 69.30it/s, 31 steps of size 2.11e-01. acc. prob=0.78]
warmup: 26%|██▌ | 258/1000 [00:38<00:10, 74.17it/s, 15 steps of size 4.11e-01. acc. prob=0.78]
warmup: 26%|██▋ | 264/1000 [00:39<00:10, 69.91it/s, 31 steps of size 2.75e-01. acc. prob=0.78]
warmup: 27%|██▋ | 271/1000 [00:39<00:10, 69.19it/s, 15 steps of size 8.86e-02. acc. prob=0.78]
warmup: 28%|██▊ | 277/1000 [00:39<00:12, 59.90it/s, 63 steps of size 9.69e-02. acc. prob=0.78]
warmup: 28%|██▊ | 284/1000 [00:39<00:11, 61.99it/s, 7 steps of size 7.24e-02. acc. prob=0.78]
warmup: 29%|██▉ | 290/1000 [00:39<00:11, 61.06it/s, 15 steps of size 9.38e-02. acc. prob=0.78]
warmup: 30%|██▉ | 297/1000 [00:39<00:11, 63.23it/s, 31 steps of size 2.99e-01. acc. prob=0.78]
warmup: 30%|███ | 304/1000 [00:39<00:10, 64.86it/s, 31 steps of size 2.46e-01. acc. prob=0.78]
warmup: 31%|███▏ | 313/1000 [00:39<00:10, 68.54it/s, 63 steps of size 1.34e-01. acc. prob=0.78]
warmup: 32%|███▏ | 321/1000 [00:39<00:10, 66.78it/s, 63 steps of size 1.22e-01. acc. prob=0.78]
warmup: 33%|███▎ | 330/1000 [00:40<00:09, 69.97it/s, 31 steps of size 2.52e-01. acc. prob=0.78]
warmup: 34%|███▎ | 336/1000 [00:40<00:09, 66.92it/s, 31 steps of size 2.36e-01. acc. prob=0.78]
warmup: 34%|███▍ | 343/1000 [00:40<00:09, 65.95it/s, 31 steps of size 2.62e-01. acc. prob=0.78]
warmup: 35%|███▌ | 353/1000 [00:40<00:08, 72.00it/s, 31 steps of size 1.34e-01. acc. prob=0.78]
warmup: 36%|███▌ | 360/1000 [00:40<00:09, 70.08it/s, 31 steps of size 2.56e-01. acc. prob=0.78]
warmup: 37%|███▋ | 370/1000 [00:40<00:08, 74.87it/s, 31 steps of size 2.29e-01. acc. prob=0.78]
warmup: 38%|███▊ | 377/1000 [00:40<00:09, 68.91it/s, 63 steps of size 1.33e-01. acc. prob=0.78]
warmup: 38%|███▊ | 385/1000 [00:40<00:08, 70.89it/s, 15 steps of size 1.26e-01. acc. prob=0.78]
warmup: 39%|███▉ | 393/1000 [00:40<00:08, 73.37it/s, 15 steps of size 2.27e-01. acc. prob=0.78]
warmup: 40%|████ | 403/1000 [00:41<00:07, 79.25it/s, 15 steps of size 1.90e-01. acc. prob=0.78]
warmup: 41%|████ | 411/1000 [00:41<00:07, 77.14it/s, 31 steps of size 2.44e-01. acc. prob=0.78]
warmup: 42%|████▏ | 420/1000 [00:41<00:07, 80.23it/s, 31 steps of size 1.65e-01. acc. prob=0.78]
warmup: 43%|████▎ | 429/1000 [00:41<00:07, 81.02it/s, 15 steps of size 2.70e-01. acc. prob=0.79]
warmup: 44%|████▎ | 437/1000 [00:41<00:06, 80.64it/s, 15 steps of size 3.90e-01. acc. prob=0.79]
warmup: 44%|████▍ | 443/1000 [00:41<00:07, 74.65it/s, 15 steps of size 2.40e-01. acc. prob=0.79]
warmup: 45%|████▌ | 454/1000 [00:41<00:07, 77.98it/s, 63 steps of size 1.14e-01. acc. prob=0.78]
warmup: 46%|████▌ | 462/1000 [00:41<00:08, 66.09it/s, 127 steps of size 7.67e-02. acc. prob=0.78]
warmup: 47%|████▋ | 469/1000 [00:41<00:08, 64.30it/s, 31 steps of size 1.92e-01. acc. prob=0.78]
warmup: 48%|████▊ | 476/1000 [00:42<00:08, 61.45it/s, 63 steps of size 8.91e-02. acc. prob=0.78]
warmup: 48%|████▊ | 483/1000 [00:42<00:08, 61.80it/s, 31 steps of size 2.48e-01. acc. prob=0.79]
warmup: 49%|████▉ | 490/1000 [00:42<00:07, 63.87it/s, 15 steps of size 1.08e-01. acc. prob=0.78]
warmup: 50%|████▉ | 497/1000 [00:42<00:07, 63.24it/s, 31 steps of size 2.43e-01. acc. prob=0.79]
sample: 50%|█████ | 505/1000 [00:42<00:07, 67.08it/s, 31 steps of size 1.86e-01. acc. prob=0.93]
sample: 51%|█████ | 512/1000 [00:42<00:07, 67.56it/s, 15 steps of size 1.86e-01. acc. prob=0.86]
sample: 52%|█████▏ | 520/1000 [00:42<00:06, 69.45it/s, 31 steps of size 1.86e-01. acc. prob=0.89]
sample: 53%|█████▎ | 527/1000 [00:42<00:06, 67.99it/s, 31 steps of size 1.86e-01. acc. prob=0.89]
sample: 53%|█████▎ | 534/1000 [00:42<00:06, 66.74it/s, 31 steps of size 1.86e-01. acc. prob=0.88]
sample: 54%|█████▍ | 542/1000 [00:43<00:06, 68.52it/s, 31 steps of size 1.86e-01. acc. prob=0.89]
sample: 55%|█████▍ | 549/1000 [00:43<00:06, 67.25it/s, 15 steps of size 1.86e-01. acc. prob=0.89]
sample: 56%|█████▌ | 556/1000 [00:43<00:06, 65.97it/s, 31 steps of size 1.86e-01. acc. prob=0.89]
sample: 56%|█████▋ | 564/1000 [00:43<00:06, 66.45it/s, 31 steps of size 1.86e-01. acc. prob=0.89]
sample: 57%|█████▋ | 570/1000 [00:43<00:06, 62.82it/s, 31 steps of size 1.86e-01. acc. prob=0.89]
sample: 58%|█████▊ | 576/1000 [00:43<00:06, 61.77it/s, 31 steps of size 1.86e-01. acc. prob=0.89]
sample: 58%|█████▊ | 582/1000 [00:43<00:06, 61.11it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 59%|█████▉ | 588/1000 [00:43<00:06, 60.65it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 59%|█████▉ | 594/1000 [00:43<00:06, 58.99it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 60%|██████ | 601/1000 [00:44<00:06, 60.74it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 61%|██████ | 609/1000 [00:44<00:05, 66.15it/s, 15 steps of size 1.86e-01. acc. prob=0.90]
sample: 62%|██████▏ | 619/1000 [00:44<00:05, 74.56it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 62%|██████▎ | 625/1000 [00:44<00:05, 69.26it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 63%|██████▎ | 632/1000 [00:44<00:05, 66.70it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 64%|██████▍ | 640/1000 [00:44<00:05, 68.82it/s, 15 steps of size 1.86e-01. acc. prob=0.90]
sample: 65%|██████▍ | 647/1000 [00:44<00:05, 66.36it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 65%|██████▌ | 653/1000 [00:44<00:05, 62.70it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 66%|██████▌ | 661/1000 [00:44<00:05, 66.31it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 67%|██████▋ | 668/1000 [00:45<00:04, 66.66it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 68%|██████▊ | 677/1000 [00:45<00:04, 69.98it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 68%|██████▊ | 685/1000 [00:45<00:04, 69.24it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 69%|██████▉ | 693/1000 [00:45<00:04, 68.89it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 70%|███████ | 701/1000 [00:45<00:04, 70.26it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 71%|███████ | 709/1000 [00:45<00:04, 71.46it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 72%|███████▏ | 715/1000 [00:45<00:04, 68.05it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 72%|███████▏ | 723/1000 [00:45<00:03, 69.71it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 73%|███████▎ | 728/1000 [00:45<00:04, 63.92it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 74%|███████▎ | 737/1000 [00:46<00:03, 68.08it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 74%|███████▍ | 745/1000 [00:46<00:03, 68.09it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 75%|███████▌ | 752/1000 [00:46<00:03, 65.51it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 76%|███████▌ | 759/1000 [00:46<00:03, 65.07it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 77%|███████▋ | 766/1000 [00:46<00:03, 65.06it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 77%|███████▋ | 773/1000 [00:46<00:03, 66.38it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 78%|███████▊ | 779/1000 [00:46<00:03, 64.28it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 79%|███████▊ | 786/1000 [00:46<00:03, 63.89it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 79%|███████▉ | 793/1000 [00:46<00:03, 63.23it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 80%|███████▉ | 799/1000 [00:47<00:03, 60.47it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 81%|████████ | 806/1000 [00:47<00:03, 62.07it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 82%|████████▏ | 815/1000 [00:47<00:02, 66.92it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 82%|████████▏ | 824/1000 [00:47<00:02, 72.19it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 83%|████████▎ | 832/1000 [00:47<00:02, 72.73it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 84%|████████▍ | 839/1000 [00:47<00:02, 70.33it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
sample: 85%|████████▍ | 846/1000 [00:47<00:02, 70.17it/s, 15 steps of size 1.86e-01. acc. prob=0.90]
sample: 85%|████████▌ | 853/1000 [00:47<00:02, 68.19it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 86%|████████▌ | 860/1000 [00:47<00:02, 68.61it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 86%|████████▋ | 865/1000 [00:47<00:02, 63.05it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 87%|████████▋ | 871/1000 [00:48<00:02, 60.51it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 88%|████████▊ | 878/1000 [00:48<00:01, 61.71it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 88%|████████▊ | 885/1000 [00:48<00:01, 61.05it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 89%|████████▉ | 891/1000 [00:48<00:01, 60.64it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 90%|████████▉ | 897/1000 [00:48<00:01, 58.78it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 90%|█████████ | 904/1000 [00:48<00:01, 60.06it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 91%|█████████ | 911/1000 [00:48<00:01, 61.55it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 92%|█████████▏| 918/1000 [00:48<00:01, 62.89it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 92%|█████████▏| 924/1000 [00:48<00:01, 61.98it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 93%|█████████▎| 931/1000 [00:49<00:01, 61.35it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 94%|█████████▍| 938/1000 [00:49<00:01, 60.74it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 95%|█████████▍| 946/1000 [00:49<00:00, 64.71it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 95%|█████████▌| 952/1000 [00:49<00:00, 62.51it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 96%|█████████▌| 958/1000 [00:49<00:00, 59.19it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 96%|█████████▋| 965/1000 [00:49<00:00, 61.31it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 97%|█████████▋| 971/1000 [00:49<00:00, 59.15it/s, 31 steps of size 1.86e-01. acc. prob=0.91]
sample: 98%|█████████▊| 979/1000 [00:49<00:00, 63.73it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 98%|█████████▊| 985/1000 [00:49<00:00, 57.96it/s, 63 steps of size 1.86e-01. acc. prob=0.91]
sample: 100%|█████████▉| 995/1000 [00:50<00:00, 68.36it/s, 15 steps of size 1.86e-01. acc. prob=0.91]
sample: 100%|██████████| 1000/1000 [00:50<00:00, 19.94it/s, 31 steps of size 1.86e-01. acc. prob=0.90]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:11<03:45, 4.21it/s, 1023 steps of size 5.55e-03. acc. prob=0.74]
warmup: 5%|▌ | 50/1000 [00:29<03:45, 4.21it/s, 1023 steps of size 4.35e-03. acc. prob=0.75]
warmup: 9%|▉ | 92/1000 [00:29<05:07, 2.95it/s, 1023 steps of size 5.70e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:29<05:06, 2.96it/s, 511 steps of size 5.29e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:30<05:12, 2.90it/s, 1023 steps of size 6.44e-03. acc. prob=0.76]
warmup: 10%|▉ | 95/1000 [00:30<05:10, 2.92it/s, 511 steps of size 6.54e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:31<05:07, 2.94it/s, 511 steps of size 4.06e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:31<05:21, 2.81it/s, 1023 steps of size 4.25e-03. acc. prob=0.76]
warmup: 10%|▉ | 98/1000 [00:32<05:16, 2.85it/s, 511 steps of size 6.29e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:32<05:10, 2.90it/s, 511 steps of size 5.61e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:32<05:04, 2.96it/s, 511 steps of size 8.31e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:32<04:57, 3.02it/s, 511 steps of size 1.19e-01. acc. prob=0.76]
warmup: 11%|█ | 106/1000 [00:33<02:28, 6.03it/s, 31 steps of size 1.24e-01. acc. prob=0.76]
warmup: 11%|█ | 112/1000 [00:33<01:24, 10.48it/s, 15 steps of size 1.39e-01. acc. prob=0.77]
warmup: 12%|█▏ | 117/1000 [00:33<01:00, 14.63it/s, 63 steps of size 1.32e-01. acc. prob=0.77]
warmup: 12%|█▏ | 124/1000 [00:33<00:40, 21.81it/s, 63 steps of size 1.23e-01. acc. prob=0.77]
warmup: 13%|█▎ | 130/1000 [00:33<00:33, 25.95it/s, 127 steps of size 6.52e-02. acc. prob=0.77]
warmup: 14%|█▎ | 136/1000 [00:33<00:27, 31.69it/s, 4 steps of size 4.35e-02. acc. prob=0.77]
warmup: 14%|█▍ | 138/1000 [00:33<00:29, 28.81it/s, 63 steps of size 1.18e-01. acc. prob=0.77]
warmup: 14%|█▍ | 145/1000 [00:33<00:22, 37.21it/s, 15 steps of size 4.85e-02. acc. prob=0.77]
warmup: 15%|█▍ | 147/1000 [00:33<00:26, 32.59it/s, 63 steps of size 1.21e-01. acc. prob=0.77]
warmup: 15%|█▌ | 154/1000 [00:34<00:21, 39.04it/s, 63 steps of size 7.85e-02. acc. prob=0.77]
warmup: 16%|█▌ | 157/1000 [00:34<00:23, 36.65it/s, 15 steps of size 1.78e-01. acc. prob=0.77]
warmup: 16%|█▌ | 161/1000 [00:34<00:24, 33.89it/s, 63 steps of size 1.36e-01. acc. prob=0.77]
warmup: 16%|█▋ | 165/1000 [00:34<00:24, 33.95it/s, 127 steps of size 6.78e-02. acc. prob=0.77]
warmup: 17%|█▋ | 171/1000 [00:34<00:24, 34.13it/s, 127 steps of size 7.48e-02. acc. prob=0.77]
warmup: 18%|█▊ | 176/1000 [00:34<00:23, 34.99it/s, 63 steps of size 1.19e-01. acc. prob=0.77]
warmup: 18%|█▊ | 182/1000 [00:34<00:20, 40.85it/s, 31 steps of size 1.86e-01. acc. prob=0.78]
warmup: 19%|█▉ | 189/1000 [00:34<00:17, 47.01it/s, 31 steps of size 1.56e-01. acc. prob=0.78]
warmup: 20%|█▉ | 197/1000 [00:35<00:15, 53.11it/s, 31 steps of size 2.31e-01. acc. prob=0.78]
warmup: 20%|██ | 202/1000 [00:35<00:15, 50.82it/s, 31 steps of size 2.71e-01. acc. prob=0.78]
warmup: 21%|██ | 211/1000 [00:35<00:13, 59.95it/s, 15 steps of size 2.71e-01. acc. prob=0.78]
warmup: 22%|██▏ | 217/1000 [00:35<00:13, 59.45it/s, 31 steps of size 3.08e-01. acc. prob=0.78]
warmup: 22%|██▎ | 225/1000 [00:35<00:12, 64.45it/s, 63 steps of size 1.29e-01. acc. prob=0.78]
warmup: 23%|██▎ | 232/1000 [00:35<00:12, 63.13it/s, 31 steps of size 2.49e-01. acc. prob=0.78]
warmup: 24%|██▍ | 240/1000 [00:35<00:11, 65.34it/s, 31 steps of size 2.19e-01. acc. prob=0.78]
warmup: 25%|██▍ | 249/1000 [00:35<00:10, 70.60it/s, 15 steps of size 2.55e-01. acc. prob=0.78]
warmup: 26%|██▌ | 255/1000 [00:35<00:11, 63.13it/s, 63 steps of size 8.98e-02. acc. prob=0.78]
warmup: 26%|██▌ | 258/1000 [00:36<00:16, 44.45it/s, 255 steps of size 2.30e-02. acc. prob=0.78]
warmup: 26%|██▌ | 260/1000 [00:36<00:21, 34.57it/s, 127 steps of size 7.21e-02. acc. prob=0.78]
warmup: 26%|██▋ | 265/1000 [00:36<00:24, 29.66it/s, 255 steps of size 4.10e-02. acc. prob=0.78]
warmup: 27%|██▋ | 267/1000 [00:36<00:27, 27.12it/s, 63 steps of size 1.44e-01. acc. prob=0.78]
warmup: 27%|██▋ | 273/1000 [00:36<00:21, 33.76it/s, 31 steps of size 1.57e-01. acc. prob=0.78]
warmup: 28%|██▊ | 277/1000 [00:36<00:21, 33.46it/s, 63 steps of size 1.37e-01. acc. prob=0.78]
warmup: 28%|██▊ | 282/1000 [00:36<00:19, 37.47it/s, 15 steps of size 8.79e-02. acc. prob=0.78]
warmup: 29%|██▊ | 287/1000 [00:37<00:19, 36.35it/s, 127 steps of size 6.15e-02. acc. prob=0.78]
warmup: 29%|██▉ | 291/1000 [00:37<00:19, 37.18it/s, 15 steps of size 7.06e-02. acc. prob=0.78]
warmup: 30%|██▉ | 295/1000 [00:37<00:19, 37.03it/s, 63 steps of size 8.72e-02. acc. prob=0.78]
warmup: 30%|███ | 301/1000 [00:37<00:16, 43.14it/s, 15 steps of size 2.34e-01. acc. prob=0.78]
warmup: 31%|███ | 308/1000 [00:37<00:13, 50.41it/s, 15 steps of size 1.86e-01. acc. prob=0.78]
warmup: 31%|███▏ | 313/1000 [00:37<00:14, 48.61it/s, 31 steps of size 8.99e-02. acc. prob=0.78]
warmup: 32%|███▏ | 317/1000 [00:37<00:15, 45.00it/s, 31 steps of size 1.44e-01. acc. prob=0.78]
warmup: 32%|███▏ | 323/1000 [00:37<00:13, 48.88it/s, 31 steps of size 1.75e-01. acc. prob=0.78]
warmup: 33%|███▎ | 328/1000 [00:37<00:13, 48.06it/s, 31 steps of size 2.72e-01. acc. prob=0.78]
warmup: 33%|███▎ | 334/1000 [00:38<00:13, 50.01it/s, 31 steps of size 1.07e-01. acc. prob=0.78]
warmup: 34%|███▍ | 341/1000 [00:38<00:12, 54.59it/s, 31 steps of size 1.97e-01. acc. prob=0.78]
warmup: 35%|███▍ | 346/1000 [00:38<00:12, 51.85it/s, 31 steps of size 1.79e-01. acc. prob=0.78]
warmup: 35%|███▌ | 350/1000 [00:38<00:14, 45.09it/s, 127 steps of size 6.31e-02. acc. prob=0.78]
warmup: 35%|███▌ | 354/1000 [00:38<00:15, 42.49it/s, 31 steps of size 1.86e-01. acc. prob=0.78]
warmup: 36%|███▌ | 360/1000 [00:38<00:13, 46.48it/s, 31 steps of size 2.34e-01. acc. prob=0.78]
warmup: 36%|███▋ | 364/1000 [00:38<00:14, 42.97it/s, 31 steps of size 1.87e-01. acc. prob=0.78]
warmup: 37%|███▋ | 369/1000 [00:38<00:14, 43.89it/s, 31 steps of size 1.21e-01. acc. prob=0.78]
warmup: 37%|███▋ | 374/1000 [00:38<00:13, 44.76it/s, 31 steps of size 2.13e-01. acc. prob=0.78]
warmup: 38%|███▊ | 380/1000 [00:39<00:13, 46.64it/s, 31 steps of size 1.69e-01. acc. prob=0.78]
warmup: 39%|███▊ | 386/1000 [00:39<00:12, 50.25it/s, 7 steps of size 1.04e-01. acc. prob=0.78]
warmup: 39%|███▉ | 391/1000 [00:39<00:12, 48.47it/s, 31 steps of size 8.50e-02. acc. prob=0.78]
warmup: 40%|███▉ | 396/1000 [00:39<00:12, 47.38it/s, 63 steps of size 7.99e-02. acc. prob=0.78]
warmup: 40%|████ | 402/1000 [00:39<00:11, 50.80it/s, 31 steps of size 5.68e-02. acc. prob=0.78]
warmup: 41%|████ | 407/1000 [00:39<00:11, 49.46it/s, 31 steps of size 1.64e-01. acc. prob=0.78]
warmup: 41%|████ | 412/1000 [00:39<00:12, 48.14it/s, 63 steps of size 8.27e-02. acc. prob=0.78]
warmup: 42%|████▏ | 417/1000 [00:39<00:12, 48.45it/s, 15 steps of size 1.21e-01. acc. prob=0.78]
warmup: 42%|████▏ | 423/1000 [00:39<00:11, 49.36it/s, 31 steps of size 1.52e-01. acc. prob=0.78]
warmup: 43%|████▎ | 428/1000 [00:39<00:11, 48.18it/s, 31 steps of size 1.60e-01. acc. prob=0.78]
warmup: 44%|████▎ | 435/1000 [00:40<00:10, 52.93it/s, 31 steps of size 1.17e-01. acc. prob=0.78]
warmup: 44%|████▍ | 441/1000 [00:40<00:10, 53.26it/s, 31 steps of size 1.58e-01. acc. prob=0.79]
warmup: 45%|████▍ | 447/1000 [00:40<00:10, 54.22it/s, 63 steps of size 9.26e-02. acc. prob=0.78]
warmup: 45%|████▌ | 450/1000 [00:40<00:11, 45.96it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 46%|████▌ | 460/1000 [00:40<00:09, 59.17it/s, 31 steps of size 1.55e-01. acc. prob=0.78]
warmup: 46%|████▋ | 463/1000 [00:40<00:10, 49.15it/s, 127 steps of size 8.39e-02. acc. prob=0.78]
warmup: 47%|████▋ | 469/1000 [00:40<00:11, 45.02it/s, 127 steps of size 6.96e-02. acc. prob=0.78]
warmup: 48%|████▊ | 475/1000 [00:40<00:10, 47.95it/s, 31 steps of size 1.96e-01. acc. prob=0.78]
warmup: 48%|████▊ | 482/1000 [00:41<00:10, 50.59it/s, 63 steps of size 1.27e-01. acc. prob=0.78]
warmup: 49%|████▉ | 490/1000 [00:41<00:09, 55.98it/s, 63 steps of size 1.31e-01. acc. prob=0.78]
warmup: 50%|████▉ | 497/1000 [00:41<00:09, 55.68it/s, 63 steps of size 1.37e-01. acc. prob=0.78]
sample: 50%|█████ | 504/1000 [00:41<00:08, 58.96it/s, 31 steps of size 1.81e-01. acc. prob=0.94]
sample: 51%|█████ | 510/1000 [00:41<00:08, 57.84it/s, 31 steps of size 1.81e-01. acc. prob=0.96]
sample: 52%|█████▏ | 517/1000 [00:41<00:08, 58.38it/s, 31 steps of size 1.81e-01. acc. prob=0.94]
sample: 52%|█████▎ | 525/1000 [00:41<00:07, 61.32it/s, 31 steps of size 1.81e-01. acc. prob=0.93]
sample: 53%|█████▎ | 532/1000 [00:41<00:07, 62.32it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 54%|█████▍ | 538/1000 [00:41<00:07, 59.86it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 55%|█████▍ | 545/1000 [00:42<00:07, 61.21it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 55%|█████▌ | 551/1000 [00:42<00:07, 59.28it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 56%|█████▌ | 557/1000 [00:42<00:07, 57.74it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 56%|█████▋ | 563/1000 [00:42<00:07, 58.15it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 57%|█████▋ | 570/1000 [00:42<00:07, 60.51it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 58%|█████▊ | 576/1000 [00:42<00:07, 60.01it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 58%|█████▊ | 582/1000 [00:42<00:07, 57.94it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 59%|█████▉ | 588/1000 [00:42<00:07, 58.35it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 60%|█████▉ | 595/1000 [00:42<00:06, 58.76it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 60%|██████ | 602/1000 [00:43<00:06, 60.24it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 61%|██████ | 608/1000 [00:43<00:06, 59.86it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 61%|██████▏ | 614/1000 [00:43<00:06, 57.95it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 62%|██████▏ | 620/1000 [00:43<00:06, 56.97it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 63%|██████▎ | 627/1000 [00:43<00:06, 59.03it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 63%|██████▎ | 634/1000 [00:43<00:06, 59.15it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 64%|██████▍ | 640/1000 [00:43<00:06, 57.58it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 65%|██████▍ | 646/1000 [00:43<00:06, 57.54it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 65%|██████▌ | 654/1000 [00:43<00:05, 60.45it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 66%|██████▌ | 660/1000 [00:44<00:05, 59.87it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 67%|██████▋ | 666/1000 [00:44<00:05, 58.26it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 67%|██████▋ | 672/1000 [00:44<00:05, 56.72it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 68%|██████▊ | 677/1000 [00:44<00:05, 54.53it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 68%|██████▊ | 683/1000 [00:44<00:05, 53.97it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 69%|██████▉ | 691/1000 [00:44<00:05, 59.29it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 70%|██████▉ | 697/1000 [00:44<00:05, 59.42it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 70%|███████ | 703/1000 [00:44<00:05, 57.89it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 71%|███████ | 710/1000 [00:44<00:04, 60.99it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 72%|███████▏ | 717/1000 [00:44<00:04, 62.18it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 72%|███████▎ | 725/1000 [00:45<00:04, 65.62it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 73%|███████▎ | 731/1000 [00:45<00:04, 62.29it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 74%|███████▍ | 738/1000 [00:45<00:04, 62.03it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 75%|███████▍ | 746/1000 [00:45<00:04, 62.00it/s, 63 steps of size 1.81e-01. acc. prob=0.92]
sample: 75%|███████▌ | 754/1000 [00:45<00:03, 66.78it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 76%|███████▌ | 762/1000 [00:45<00:03, 68.22it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 77%|███████▋ | 770/1000 [00:45<00:03, 68.55it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 78%|███████▊ | 776/1000 [00:45<00:03, 65.98it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 78%|███████▊ | 782/1000 [00:45<00:03, 64.04it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 79%|███████▉ | 791/1000 [00:46<00:02, 69.80it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 80%|███████▉ | 798/1000 [00:46<00:02, 69.67it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 81%|████████ | 806/1000 [00:46<00:02, 69.58it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 81%|████████ | 812/1000 [00:46<00:02, 66.86it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 82%|████████▏ | 818/1000 [00:46<00:02, 63.00it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 82%|████████▎ | 825/1000 [00:46<00:02, 61.89it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 83%|████████▎ | 831/1000 [00:46<00:02, 61.27it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 84%|████████▍ | 838/1000 [00:46<00:02, 63.65it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 84%|████████▍ | 845/1000 [00:46<00:02, 64.08it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 85%|████████▌ | 853/1000 [00:47<00:02, 66.95it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 86%|████████▌ | 860/1000 [00:47<00:02, 64.71it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 87%|████████▋ | 870/1000 [00:47<00:01, 71.35it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 88%|████████▊ | 877/1000 [00:47<00:01, 70.79it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 88%|████████▊ | 884/1000 [00:47<00:01, 68.88it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 89%|████████▉ | 890/1000 [00:47<00:01, 65.96it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 90%|████████▉ | 897/1000 [00:47<00:01, 64.94it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 90%|█████████ | 905/1000 [00:47<00:01, 65.92it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 91%|█████████ | 911/1000 [00:47<00:01, 63.06it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 92%|█████████▏| 918/1000 [00:48<00:01, 63.35it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 92%|█████████▏| 924/1000 [00:48<00:01, 61.88it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 93%|█████████▎| 931/1000 [00:48<00:01, 62.52it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 94%|█████████▍| 939/1000 [00:48<00:00, 67.19it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 94%|█████████▍| 945/1000 [00:48<00:00, 63.10it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 95%|█████████▌| 952/1000 [00:48<00:00, 63.45it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 96%|█████████▌| 958/1000 [00:48<00:00, 62.46it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 97%|█████████▋| 966/1000 [00:48<00:00, 66.51it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 97%|█████████▋| 973/1000 [00:48<00:00, 65.95it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 98%|█████████▊| 981/1000 [00:48<00:00, 68.65it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 99%|█████████▉| 989/1000 [00:49<00:00, 71.65it/s, 15 steps of size 1.81e-01. acc. prob=0.92]
sample: 100%|█████████▉| 996/1000 [00:49<00:00, 67.74it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
sample: 100%|██████████| 1000/1000 [00:49<00:00, 20.29it/s, 31 steps of size 1.81e-01. acc. prob=0.92]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:12<04:03, 3.90it/s, 5 steps of size 1.41e-03. acc. prob=0.72]
warmup: 5%|▌ | 50/1000 [00:29<04:03, 3.90it/s, 511 steps of size 2.76e-03. acc. prob=0.75]
warmup: 9%|▉ | 92/1000 [00:30<05:12, 2.90it/s, 1023 steps of size 4.17e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:30<05:11, 2.91it/s, 511 steps of size 5.76e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:30<05:10, 2.92it/s, 511 steps of size 3.30e-03. acc. prob=0.75]
warmup: 10%|▉ | 95/1000 [00:31<05:17, 2.85it/s, 1023 steps of size 4.75e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:31<05:14, 2.87it/s, 511 steps of size 6.73e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:32<05:11, 2.90it/s, 511 steps of size 8.83e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:32<05:03, 2.97it/s, 1023 steps of size 3.08e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:33<05:26, 2.76it/s, 1023 steps of size 3.01e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:33<05:50, 2.57it/s, 1023 steps of size 4.33e-02. acc. prob=0.76]
warmup: 10%|█ | 103/1000 [00:34<04:30, 3.32it/s, 63 steps of size 1.48e-01. acc. prob=0.76]
warmup: 11%|█ | 109/1000 [00:34<02:15, 6.58it/s, 31 steps of size 3.40e-01. acc. prob=0.77]
warmup: 11%|█▏ | 113/1000 [00:34<01:38, 8.97it/s, 63 steps of size 1.12e-01. acc. prob=0.76]
warmup: 12%|█▏ | 120/1000 [00:34<01:01, 14.21it/s, 127 steps of size 6.41e-02. acc. prob=0.76]
warmup: 12%|█▏ | 124/1000 [00:34<00:50, 17.18it/s, 15 steps of size 4.54e-01. acc. prob=0.77]
warmup: 13%|█▎ | 127/1000 [00:34<00:46, 18.69it/s, 63 steps of size 1.34e-01. acc. prob=0.77]
warmup: 13%|█▎ | 134/1000 [00:34<00:32, 26.93it/s, 15 steps of size 2.41e-01. acc. prob=0.77]
warmup: 14%|█▍ | 140/1000 [00:34<00:26, 32.84it/s, 15 steps of size 3.71e-01. acc. prob=0.78]
warmup: 14%|█▍ | 144/1000 [00:34<00:25, 34.24it/s, 31 steps of size 1.98e-01. acc. prob=0.77]
warmup: 15%|█▌ | 151/1000 [00:35<00:20, 41.99it/s, 31 steps of size 1.96e+00. acc. prob=0.78]
warmup: 16%|█▌ | 156/1000 [00:35<00:20, 42.05it/s, 31 steps of size 1.97e-01. acc. prob=0.77]
warmup: 16%|█▋ | 163/1000 [00:35<00:18, 45.60it/s, 63 steps of size 1.15e-01. acc. prob=0.77]
warmup: 17%|█▋ | 168/1000 [00:35<00:20, 40.29it/s, 127 steps of size 5.88e-02. acc. prob=0.77]
warmup: 17%|█▋ | 173/1000 [00:35<00:21, 38.35it/s, 127 steps of size 5.80e-02. acc. prob=0.77]
warmup: 18%|█▊ | 178/1000 [00:35<00:20, 40.98it/s, 63 steps of size 1.05e-01. acc. prob=0.77]
warmup: 18%|█▊ | 182/1000 [00:35<00:20, 40.47it/s, 127 steps of size 6.21e-02. acc. prob=0.77]
warmup: 19%|█▊ | 186/1000 [00:35<00:20, 40.11it/s, 15 steps of size 2.15e-01. acc. prob=0.78]
warmup: 19%|█▉ | 191/1000 [00:35<00:18, 42.73it/s, 31 steps of size 1.90e-01. acc. prob=0.78]
warmup: 20%|█▉ | 198/1000 [00:36<00:18, 42.72it/s, 127 steps of size 6.60e-02. acc. prob=0.77]
warmup: 20%|██ | 204/1000 [00:36<00:17, 46.23it/s, 15 steps of size 3.42e-01. acc. prob=0.78]
warmup: 21%|██ | 209/1000 [00:36<00:17, 46.48it/s, 15 steps of size 2.55e-01. acc. prob=0.78]
warmup: 22%|██▏ | 215/1000 [00:36<00:16, 47.22it/s, 63 steps of size 1.98e-01. acc. prob=0.78]
warmup: 22%|██▏ | 219/1000 [00:36<00:17, 44.43it/s, 31 steps of size 1.53e-01. acc. prob=0.78]
warmup: 22%|██▏ | 224/1000 [00:36<00:18, 41.69it/s, 127 steps of size 7.30e-02. acc. prob=0.78]
warmup: 23%|██▎ | 228/1000 [00:36<00:19, 40.36it/s, 31 steps of size 2.42e-01. acc. prob=0.78]
warmup: 24%|██▎ | 237/1000 [00:36<00:15, 49.45it/s, 63 steps of size 1.71e-01. acc. prob=0.78]
warmup: 24%|██▍ | 245/1000 [00:37<00:14, 53.14it/s, 63 steps of size 7.59e-02. acc. prob=0.78]
warmup: 25%|██▌ | 253/1000 [00:37<00:13, 54.94it/s, 63 steps of size 9.60e-02. acc. prob=0.78]
warmup: 26%|██▌ | 260/1000 [00:37<00:14, 51.04it/s, 127 steps of size 6.59e-02. acc. prob=0.78]
warmup: 26%|██▋ | 265/1000 [00:37<00:14, 49.39it/s, 31 steps of size 1.03e-01. acc. prob=0.78]
warmup: 27%|██▋ | 270/1000 [00:37<00:15, 48.38it/s, 31 steps of size 1.68e-01. acc. prob=0.78]
warmup: 28%|██▊ | 276/1000 [00:37<00:14, 50.29it/s, 31 steps of size 1.21e-01. acc. prob=0.78]
warmup: 28%|██▊ | 284/1000 [00:37<00:12, 56.93it/s, 63 steps of size 7.92e-02. acc. prob=0.78]
warmup: 29%|██▉ | 289/1000 [00:37<00:13, 52.88it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 30%|██▉ | 296/1000 [00:38<00:12, 55.74it/s, 63 steps of size 7.88e-02. acc. prob=0.78]
warmup: 30%|███ | 302/1000 [00:38<00:12, 55.20it/s, 31 steps of size 1.91e-01. acc. prob=0.78]
warmup: 31%|███ | 310/1000 [00:38<00:11, 59.33it/s, 31 steps of size 2.66e-01. acc. prob=0.78]
warmup: 32%|███▏ | 316/1000 [00:38<00:11, 58.56it/s, 63 steps of size 1.24e-01. acc. prob=0.78]
warmup: 32%|███▏ | 323/1000 [00:38<00:11, 58.95it/s, 31 steps of size 1.90e-01. acc. prob=0.78]
warmup: 33%|███▎ | 328/1000 [00:38<00:11, 56.22it/s, 31 steps of size 1.67e-01. acc. prob=0.78]
warmup: 34%|███▎ | 336/1000 [00:38<00:11, 59.88it/s, 31 steps of size 1.54e-01. acc. prob=0.78]
warmup: 34%|███▍ | 344/1000 [00:38<00:10, 63.89it/s, 31 steps of size 2.25e-01. acc. prob=0.78]
warmup: 35%|███▌ | 351/1000 [00:38<00:10, 60.84it/s, 63 steps of size 1.13e-01. acc. prob=0.78]
warmup: 36%|███▌ | 356/1000 [00:39<00:11, 55.95it/s, 31 steps of size 1.82e-01. acc. prob=0.78]
warmup: 36%|███▌ | 362/1000 [00:39<00:11, 55.14it/s, 31 steps of size 1.92e-01. acc. prob=0.78]
warmup: 37%|███▋ | 371/1000 [00:39<00:09, 63.30it/s, 15 steps of size 1.12e-01. acc. prob=0.78]
warmup: 38%|███▊ | 378/1000 [00:39<00:09, 63.11it/s, 31 steps of size 2.22e-01. acc. prob=0.78]
warmup: 39%|███▊ | 386/1000 [00:39<00:09, 65.62it/s, 31 steps of size 1.43e-01. acc. prob=0.78]
warmup: 39%|███▉ | 392/1000 [00:39<00:09, 62.12it/s, 31 steps of size 1.40e-01. acc. prob=0.78]
warmup: 40%|███▉ | 399/1000 [00:39<00:09, 61.80it/s, 31 steps of size 1.98e-01. acc. prob=0.78]
warmup: 41%|████ | 408/1000 [00:39<00:09, 63.28it/s, 63 steps of size 8.23e-02. acc. prob=0.78]
warmup: 42%|████▏ | 415/1000 [00:39<00:08, 65.02it/s, 31 steps of size 1.94e-01. acc. prob=0.78]
warmup: 42%|████▏ | 423/1000 [00:40<00:08, 67.50it/s, 31 steps of size 1.96e-01. acc. prob=0.78]
warmup: 43%|████▎ | 431/1000 [00:40<00:08, 69.03it/s, 31 steps of size 1.59e-01. acc. prob=0.78]
warmup: 44%|████▍ | 439/1000 [00:40<00:07, 70.28it/s, 31 steps of size 1.48e-01. acc. prob=0.78]
warmup: 45%|████▍ | 447/1000 [00:40<00:07, 71.18it/s, 31 steps of size 1.57e-01. acc. prob=0.78]
warmup: 46%|████▌ | 456/1000 [00:40<00:07, 75.42it/s, 15 steps of size 2.56e-01. acc. prob=0.78]
warmup: 46%|████▌ | 460/1000 [00:40<00:08, 61.93it/s, 63 steps of size 1.04e-01. acc. prob=0.78]
warmup: 47%|████▋ | 467/1000 [00:40<00:09, 58.18it/s, 127 steps of size 7.17e-02. acc. prob=0.78]
warmup: 47%|████▋ | 473/1000 [00:40<00:10, 50.04it/s, 127 steps of size 8.54e-02. acc. prob=0.78]
warmup: 48%|████▊ | 479/1000 [00:41<00:10, 51.35it/s, 31 steps of size 2.09e-01. acc. prob=0.78]
warmup: 48%|████▊ | 485/1000 [00:41<00:09, 52.90it/s, 31 steps of size 1.78e-01. acc. prob=0.78]
warmup: 49%|████▉ | 491/1000 [00:41<00:09, 53.78it/s, 31 steps of size 2.68e-01. acc. prob=0.78]
warmup: 50%|████▉ | 498/1000 [00:41<00:08, 57.10it/s, 15 steps of size 2.17e-01. acc. prob=0.78]
sample: 50%|█████ | 505/1000 [00:41<00:08, 60.51it/s, 15 steps of size 1.93e-01. acc. prob=0.95]
sample: 51%|█████ | 512/1000 [00:41<00:07, 63.15it/s, 15 steps of size 1.93e-01. acc. prob=0.86]
sample: 52%|█████▏ | 520/1000 [00:41<00:07, 67.86it/s, 31 steps of size 1.93e-01. acc. prob=0.85]
sample: 53%|█████▎ | 528/1000 [00:41<00:06, 69.76it/s, 15 steps of size 1.93e-01. acc. prob=0.87]
sample: 54%|█████▎ | 535/1000 [00:41<00:07, 66.42it/s, 31 steps of size 1.93e-01. acc. prob=0.88]
sample: 54%|█████▍ | 542/1000 [00:42<00:06, 65.91it/s, 31 steps of size 1.93e-01. acc. prob=0.90]
sample: 55%|█████▌ | 550/1000 [00:42<00:06, 69.79it/s, 15 steps of size 1.93e-01. acc. prob=0.89]
sample: 56%|█████▌ | 555/1000 [00:42<00:06, 63.84it/s, 31 steps of size 1.93e-01. acc. prob=0.89]
sample: 56%|█████▌ | 561/1000 [00:42<00:07, 62.63it/s, 31 steps of size 1.93e-01. acc. prob=0.89]
sample: 57%|█████▋ | 568/1000 [00:42<00:06, 61.99it/s, 31 steps of size 1.93e-01. acc. prob=0.90]
sample: 57%|█████▊ | 575/1000 [00:42<00:06, 61.85it/s, 31 steps of size 1.93e-01. acc. prob=0.90]
sample: 58%|█████▊ | 583/1000 [00:42<00:06, 65.49it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 59%|█████▉ | 589/1000 [00:42<00:06, 61.92it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 60%|█████▉ | 596/1000 [00:42<00:06, 62.60it/s, 15 steps of size 1.93e-01. acc. prob=0.91]
sample: 60%|██████ | 605/1000 [00:42<00:05, 66.86it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 61%|██████ | 612/1000 [00:43<00:05, 66.53it/s, 15 steps of size 1.93e-01. acc. prob=0.92]
sample: 62%|██████▏ | 618/1000 [00:43<00:05, 64.53it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 63%|██████▎ | 626/1000 [00:43<00:05, 65.63it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 63%|██████▎ | 632/1000 [00:43<00:05, 62.50it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 64%|██████▍ | 640/1000 [00:43<00:05, 64.20it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 65%|██████▍ | 647/1000 [00:43<00:05, 65.63it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 65%|██████▌ | 653/1000 [00:43<00:05, 62.57it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 66%|██████▌ | 661/1000 [00:43<00:05, 64.61it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 67%|██████▋ | 668/1000 [00:43<00:05, 64.43it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 67%|██████▋ | 674/1000 [00:44<00:05, 61.15it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 68%|██████▊ | 681/1000 [00:44<00:05, 61.59it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 69%|██████▊ | 687/1000 [00:44<00:05, 58.34it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 69%|██████▉ | 694/1000 [00:44<00:05, 60.47it/s, 15 steps of size 1.93e-01. acc. prob=0.92]
sample: 70%|███████ | 701/1000 [00:44<00:04, 61.22it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 71%|███████ | 707/1000 [00:44<00:04, 60.70it/s, 15 steps of size 1.93e-01. acc. prob=0.92]
sample: 71%|███████ | 712/1000 [00:44<00:05, 56.66it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 72%|███████▏ | 720/1000 [00:44<00:04, 60.33it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 73%|███████▎ | 728/1000 [00:44<00:04, 64.14it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 73%|███████▎ | 734/1000 [00:45<00:04, 62.35it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 74%|███████▍ | 740/1000 [00:45<00:04, 61.58it/s, 15 steps of size 1.93e-01. acc. prob=0.92]
sample: 75%|███████▍ | 747/1000 [00:45<00:04, 62.50it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 76%|███████▌ | 755/1000 [00:45<00:03, 64.61it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 76%|███████▌ | 762/1000 [00:45<00:03, 65.53it/s, 15 steps of size 1.93e-01. acc. prob=0.92]
sample: 77%|███████▋ | 769/1000 [00:45<00:03, 66.77it/s, 15 steps of size 1.93e-01. acc. prob=0.92]
sample: 78%|███████▊ | 776/1000 [00:45<00:03, 67.67it/s, 15 steps of size 1.93e-01. acc. prob=0.92]
sample: 78%|███████▊ | 784/1000 [00:45<00:03, 67.87it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 79%|███████▉ | 793/1000 [00:45<00:02, 72.50it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 80%|████████ | 800/1000 [00:46<00:02, 70.01it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 81%|████████ | 807/1000 [00:46<00:02, 68.40it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 82%|████████▏ | 815/1000 [00:46<00:02, 70.46it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 82%|████████▏ | 822/1000 [00:46<00:02, 68.50it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 83%|████████▎ | 829/1000 [00:46<00:02, 66.95it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 84%|████████▎ | 835/1000 [00:46<00:02, 64.92it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 84%|████████▍ | 843/1000 [00:46<00:02, 67.48it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 85%|████████▌ | 850/1000 [00:46<00:02, 67.80it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 86%|████████▌ | 856/1000 [00:46<00:02, 63.99it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 86%|████████▋ | 863/1000 [00:47<00:02, 63.01it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 87%|████████▋ | 870/1000 [00:47<00:02, 62.00it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 88%|████████▊ | 877/1000 [00:47<00:01, 63.74it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 88%|████████▊ | 883/1000 [00:47<00:01, 62.44it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 89%|████████▉ | 890/1000 [00:47<00:01, 62.84it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 90%|████████▉ | 897/1000 [00:47<00:01, 61.77it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 90%|█████████ | 904/1000 [00:47<00:01, 61.00it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 91%|█████████ | 910/1000 [00:47<00:01, 60.39it/s, 31 steps of size 1.93e-01. acc. prob=0.92]
sample: 92%|█████████▏| 917/1000 [00:47<00:01, 61.54it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 92%|█████████▏| 924/1000 [00:47<00:01, 63.82it/s, 15 steps of size 1.93e-01. acc. prob=0.91]
sample: 93%|█████████▎| 930/1000 [00:48<00:01, 62.58it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 94%|█████████▎| 936/1000 [00:48<00:01, 60.38it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 94%|█████████▍| 943/1000 [00:48<00:00, 61.73it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 95%|█████████▌| 950/1000 [00:48<00:00, 62.94it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 96%|█████████▌| 957/1000 [00:48<00:00, 63.34it/s, 47 steps of size 1.93e-01. acc. prob=0.91]
sample: 97%|█████████▋| 967/1000 [00:48<00:00, 70.06it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 97%|█████████▋| 974/1000 [00:48<00:00, 68.11it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 98%|█████████▊| 981/1000 [00:48<00:00, 65.38it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 99%|█████████▉| 989/1000 [00:48<00:00, 66.51it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 100%|█████████▉| 996/1000 [00:49<00:00, 65.14it/s, 31 steps of size 1.93e-01. acc. prob=0.91]
sample: 100%|██████████| 1000/1000 [00:49<00:00, 20.35it/s, 15 steps of size 1.93e-01. acc. prob=0.91]
We recommend running at least 4 chains for robust computation of convergence diagnostics
Final notes:¶
- whether centered is better depends on the structure of your dataset
- the two models are mathematically equivalent, however the posterior geometry is affected by the parameterization
- in practice, you can try both and see which one works better
- here is a great (albeit non-trivial to digest) long blog post which touches on the topic in some detail`
Step 5: Model Validation¶
In [19]:
Copied!
#### posterior statistics of centered model
az.summary(samples_model_reg_v_ex2_A2, var_names=["~_id"], filter_vars="like")
#### posterior statistics of centered model
az.summary(samples_model_reg_v_ex2_A2, var_names=["~_id"], filter_vars="like")
Out[19]:
| mean | sd | eti89_lb | eti89_ub | ess_bulk | ess_tail | r_hat | mcse_mean | mcse_sd | |
|---|---|---|---|---|---|---|---|---|---|
| v_zGPe | 0.336 | 0.0145 | 0.31 | 0.36 | 1348 | 826 | 1.00 | 0.00044 | 0.0004 |
| z | 0.49 | 0.0113 | 0.47 | 0.51 | 1718 | 1166 | 1.00 | 0.00027 | 0.00019 |
| v_zSTN | 0.812 | 0.0188 | 0.78 | 0.84 | 1071 | 874 | 1.00 | 0.00058 | 0.00043 |
| v_Intercept | 1.56 | 0.053 | 1.5 | 1.6 | 968 | 812 | 1.00 | 0.0017 | 0.0014 |
| t | 0.5041 | 0.0057 | 0.49 | 0.51 | 1654 | 1217 | 1.01 | 0.00014 | 9.6e-05 |
| a | 1.48 | 0.0262 | 1.4 | 1.5 | 1715 | 1250 | 1.00 | 0.00063 | 0.00049 |
In [20]:
Copied!
# Plotting the posterior predictive
hssm.plotting.plot_predictive(model_reg_v_ex2_A1, col="participant_id", col_wrap=5)
# Plotting the posterior predictive
hssm.plotting.plot_predictive(model_reg_v_ex2_A1, col="participant_id", col_wrap=5)
No posterior_predictive samples found. Generating posterior_predictive samples using the provided DataTree object and the original data. This will modify the provided DataTree object, or if not provided, the traces object stored inside the model.
Out[20]:
<seaborn.axisgrid.FacetGrid at 0x142d882f0>
In [21]:
Copied!
#### posterior statistics of non-centered model
az.summary(samples_model_reg_v_ex2_A2, var_names=["~_offset"], filter_vars="like")
#### posterior statistics of non-centered model
az.summary(samples_model_reg_v_ex2_A2, var_names=["~_offset"], filter_vars="like")
Out[21]:
| mean | sd | eti89_lb | eti89_ub | ess_bulk | ess_tail | r_hat | mcse_mean | mcse_sd | |
|---|---|---|---|---|---|---|---|---|---|
| v_1|participant_id_sigma | 0.096 | 0.045 | 0.034 | 0.17 | 494 | 626 | 1.00 | 0.002 | 0.0024 |
| v_zGPe | 0.336 | 0.0145 | 0.31 | 0.36 | 1348 | 826 | 1.00 | 0.00044 | 0.0004 |
| z | 0.49 | 0.0113 | 0.47 | 0.51 | 1718 | 1166 | 1.00 | 0.00027 | 0.00019 |
| v_zSTN|participant_id[1] | -0.022 | 0.027 | -0.074 | 0.0099 | 936 | 1269 | 1.00 | 0.0009 | 0.00073 |
| v_zSTN|participant_id[10] | -0.007 | 0.024 | -0.048 | 0.028 | 1506 | 1134 | 1.00 | 0.00063 | 0.00063 |
| v_zSTN|participant_id[2] | -0.006 | 0.025 | -0.048 | 0.032 | 1554 | 1210 | 1.00 | 0.00064 | 0.00058 |
| v_zSTN|participant_id[3] | -0.006 | 0.026 | -0.051 | 0.033 | 1443 | 1062 | 1.00 | 0.0007 | 0.00069 |
| v_zSTN|participant_id[4] | 0.025 | 0.03 | -0.0085 | 0.084 | 695 | 880 | 1.00 | 0.0012 | 0.00097 |
| v_zSTN|participant_id[5] | -0.003 | 0.024 | -0.041 | 0.035 | 1093 | 978 | 1.01 | 0.00073 | 0.00067 |
| v_zSTN|participant_id[6] | -0.004 | 0.023 | -0.042 | 0.031 | 1602 | 1173 | 1.00 | 0.00057 | 0.00054 |
| v_zSTN|participant_id[7] | -0.011 | 0.026 | -0.057 | 0.027 | 1486 | 1217 | 1.00 | 0.00067 | 0.00058 |
| v_zSTN|participant_id[8] | 0.022 | 0.029 | -0.013 | 0.075 | 941 | 978 | 1.00 | 0.00097 | 0.00083 |
| v_zSTN|participant_id[9] | 0.011 | 0.025 | -0.022 | 0.055 | 1421 | 958 | 1.00 | 0.00071 | 0.00076 |
| v_zSTN | 0.812 | 0.0188 | 0.78 | 0.84 | 1071 | 874 | 1.00 | 0.00058 | 0.00043 |
| v_zGPe|participant_id[1] | -0.005 | 0.02 | -0.039 | 0.022 | 1486 | 1231 | 1.00 | 0.00054 | 0.00073 |
| v_zGPe|participant_id[10] | 0.016 | 0.023 | -0.0088 | 0.062 | 973 | 1232 | 1.00 | 0.0008 | 0.00086 |
| v_zGPe|participant_id[2] | 0.0029 | 0.019 | -0.026 | 0.036 | 1764 | 1183 | 1.00 | 0.00048 | 0.00051 |
| v_zGPe|participant_id[3] | -0.009 | 0.021 | -0.048 | 0.016 | 1129 | 1077 | 1.00 | 0.00068 | 0.00078 |
| v_zGPe|participant_id[4] | 0.004 | 0.019 | -0.025 | 0.036 | 1485 | 1111 | 1.00 | 0.00052 | 0.00056 |
| v_zGPe|participant_id[5] | -0 | 0.02 | -0.033 | 0.033 | 1294 | 1207 | 1.00 | 0.00057 | 0.00068 |
| v_zGPe|participant_id[6] | -0.005 | 0.019 | -0.036 | 0.02 | 1622 | 1165 | 1.01 | 0.00052 | 0.00069 |
| v_zGPe|participant_id[7] | 0.01 | 0.021 | -0.015 | 0.047 | 1459 | 1267 | 1.00 | 0.00054 | 0.00056 |
| v_zGPe|participant_id[8] | -0.006 | 0.019 | -0.041 | 0.02 | 1366 | 1309 | 1.00 | 0.00053 | 0.00058 |
| v_zGPe|participant_id[9] | -0.009 | 0.02 | -0.046 | 0.017 | 1315 | 1230 | 1.01 | 0.00059 | 0.00062 |
| v_zSTN|participant_id_sigma | 0.029 | 0.019 | 0.0032 | 0.061 | 441 | 650 | 1.00 | 0.00092 | 0.00089 |
| v_zGPe|participant_id_sigma | 0.021 | 0.017 | 0.0017 | 0.05 | 446 | 594 | 1.01 | 0.00083 | 0.0012 |
| v_Intercept | 1.56 | 0.053 | 1.5 | 1.6 | 968 | 812 | 1.00 | 0.0017 | 0.0014 |
| t | 0.5041 | 0.0057 | 0.49 | 0.51 | 1654 | 1217 | 1.01 | 0.00014 | 9.6e-05 |
| v_1|participant_id[1] | 0.023 | 0.062 | -0.069 | 0.13 | 1098 | 973 | 1.00 | 0.0019 | 0.0016 |
| v_1|participant_id[10] | -0.031 | 0.061 | -0.13 | 0.064 | 1331 | 844 | 1.00 | 0.0017 | 0.0013 |
| v_1|participant_id[2] | 0.039 | 0.06 | -0.051 | 0.14 | 1261 | 1140 | 1.00 | 0.0017 | 0.0013 |
| v_1|participant_id[3] | 0.081 | 0.073 | -0.022 | 0.2 | 1075 | 1182 | 1.00 | 0.0022 | 0.002 |
| v_1|participant_id[4] | 0.025 | 0.063 | -0.074 | 0.13 | 1449 | 1129 | 1.00 | 0.0017 | 0.0016 |
| v_1|participant_id[5] | 0.099 | 0.072 | -0.0038 | 0.22 | 976 | 972 | 1.00 | 0.0023 | 0.002 |
| v_1|participant_id[6] | 0.006 | 0.065 | -0.091 | 0.11 | 1328 | 1172 | 1.01 | 0.0018 | 0.0019 |
| v_1|participant_id[7] | -0.087 | 0.069 | -0.21 | 0.011 | 1045 | 1069 | 1.00 | 0.0021 | 0.0016 |
| v_1|participant_id[8] | -0.069 | 0.065 | -0.18 | 0.024 | 1358 | 1371 | 1.00 | 0.0017 | 0.0014 |
| v_1|participant_id[9] | -0.055 | 0.064 | -0.16 | 0.037 | 1114 | 1119 | 1.00 | 0.0019 | 0.0015 |
| a | 1.48 | 0.0262 | 1.4 | 1.5 | 1715 | 1250 | 1.00 | 0.00063 | 0.00049 |
In [22]:
Copied!
# Plotting the posterior predictive
hssm.plotting.plot_predictive(model_reg_v_ex2_A2, col="participant_id", col_wrap=5)
# Plotting the posterior predictive
hssm.plotting.plot_predictive(model_reg_v_ex2_A2, col="participant_id", col_wrap=5)
No posterior_predictive samples found. Generating posterior_predictive samples using the provided DataTree object and the original data. This will modify the provided DataTree object, or if not provided, the traces object stored inside the model.
Out[22]:
<seaborn.axisgrid.FacetGrid at 0x139d9ad50>
Additional Case studies (between-subject & group variables)¶
Case Study 3: Participant-level hierarchy, two within-subject coefficients and one between-subject coefficient¶
- we still have the intracranial recordings
- but now you have different patients and they vary in symptom severity for Parkinson
- you wonder whether those with higher symptom severity have a lower drift rate because the modulation from STN is less efficient
Step 1: Regression-based data simulation (multiple subjects)¶
In [23]:
Copied!
# Function to simulate data for one participant
def simulate_participant2(participant_id, sevScore, size=300):
"""Simulate DDM trial data for one participant with severity-dependent drift."""
intercept = 0.5
zSTN = np.random.normal(loc=1, scale=2, size=size)
zGPe = np.random.normal(loc=1, scale=2, size=size)
## Strength of neural modulation depends on PD severity score
#### Case 1: only the interaction btw sevScore & STN/GPe will turn out
#### sign, the main effects of STN & GPe will disappear
# v = intercept + 0.8 * zSTN * sevScore + 0.3 * zGPe * sevScore
#### Case 2: main effects of STN/GPe will remain, plus a significant
#### interaction, but no modulation on intercept
if sevScore < 0.2:
v = intercept + 0.8 * zSTN + 0.3 * zGPe
elif sevScore >= 0.2 and sevScore < 0.4:
v = intercept + 0.7 * zSTN + 0.25 * zGPe
elif sevScore >= 0.4 and sevScore < 0.6:
v = intercept + 0.6 * zSTN + 0.2 * zGPe
elif sevScore >= 0.6 and sevScore < 0.8:
v = intercept + 0.5 * zSTN + 0.15 * zGPe
else:
v = intercept + 0.4 * zSTN + 0.1 * zGPe
# Assume `hssm.simulate_data` returns a DataFrame
true_values = np.column_stack(
[v, np.repeat([[1.5, 0.5, 0.5]], axis=0, repeats=size)]
)
dataset_reg_v = hssm.simulate_data(
model="ddm",
theta=true_values,
size=1, # Generate one data point for each of the 1000 set of true values
)
# Adding additional variables to the dataset
dataset_reg_v["zSTN"] = zSTN
dataset_reg_v["zGPe"] = zGPe
dataset_reg_v["participant_id"] = str(participant_id)
dataset_reg_v["sevScore"] = sevScore
return dataset_reg_v
# Simulate data for four participants
### note that we assume that STN & GPe
# patients with severity scores
subj_list = [1, 1, 2, 3, 4, 5, 6, 7, 8, 8, 8]
sevscore_list = [0.00, 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90, 1.00]
# Combine datasets into one DataFrame
combined_dataset2 = pd.concat(
[
simulate_participant2(subj, sevscore)
for subj, sevscore in zip(subj_list, sevscore_list)
],
ignore_index=True,
)
combined_dataset2
# Function to simulate data for one participant
def simulate_participant2(participant_id, sevScore, size=300):
"""Simulate DDM trial data for one participant with severity-dependent drift."""
intercept = 0.5
zSTN = np.random.normal(loc=1, scale=2, size=size)
zGPe = np.random.normal(loc=1, scale=2, size=size)
## Strength of neural modulation depends on PD severity score
#### Case 1: only the interaction btw sevScore & STN/GPe will turn out
#### sign, the main effects of STN & GPe will disappear
# v = intercept + 0.8 * zSTN * sevScore + 0.3 * zGPe * sevScore
#### Case 2: main effects of STN/GPe will remain, plus a significant
#### interaction, but no modulation on intercept
if sevScore < 0.2:
v = intercept + 0.8 * zSTN + 0.3 * zGPe
elif sevScore >= 0.2 and sevScore < 0.4:
v = intercept + 0.7 * zSTN + 0.25 * zGPe
elif sevScore >= 0.4 and sevScore < 0.6:
v = intercept + 0.6 * zSTN + 0.2 * zGPe
elif sevScore >= 0.6 and sevScore < 0.8:
v = intercept + 0.5 * zSTN + 0.15 * zGPe
else:
v = intercept + 0.4 * zSTN + 0.1 * zGPe
# Assume `hssm.simulate_data` returns a DataFrame
true_values = np.column_stack(
[v, np.repeat([[1.5, 0.5, 0.5]], axis=0, repeats=size)]
)
dataset_reg_v = hssm.simulate_data(
model="ddm",
theta=true_values,
size=1, # Generate one data point for each of the 1000 set of true values
)
# Adding additional variables to the dataset
dataset_reg_v["zSTN"] = zSTN
dataset_reg_v["zGPe"] = zGPe
dataset_reg_v["participant_id"] = str(participant_id)
dataset_reg_v["sevScore"] = sevScore
return dataset_reg_v
# Simulate data for four participants
### note that we assume that STN & GPe
# patients with severity scores
subj_list = [1, 1, 2, 3, 4, 5, 6, 7, 8, 8, 8]
sevscore_list = [0.00, 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90, 1.00]
# Combine datasets into one DataFrame
combined_dataset2 = pd.concat(
[
simulate_participant2(subj, sevscore)
for subj, sevscore in zip(subj_list, sevscore_list)
],
ignore_index=True,
)
combined_dataset2
Out[23]:
| rt | response | zSTN | zGPe | participant_id | sevScore | |
|---|---|---|---|---|---|---|
| 0 | 1.167871 | 1.0 | 0.542555 | 3.558076 | 1 | 0.0 |
| 1 | 0.744359 | 1.0 | 2.945089 | 1.815673 | 1 | 0.0 |
| 2 | 1.561734 | 1.0 | 0.472066 | -0.411510 | 1 | 0.0 |
| 3 | 1.042030 | 1.0 | 2.303357 | 0.950618 | 1 | 0.0 |
| 4 | 1.433222 | -1.0 | -0.543237 | -1.771657 | 1 | 0.0 |
| ... | ... | ... | ... | ... | ... | ... |
| 3295 | 1.487224 | 1.0 | 1.738488 | 2.102996 | 8 | 1.0 |
| 3296 | 1.201577 | 1.0 | 2.174505 | -0.688230 | 8 | 1.0 |
| 3297 | 3.122706 | 1.0 | 0.943154 | -1.270588 | 8 | 1.0 |
| 3298 | 1.008394 | 1.0 | 3.827790 | 2.551132 | 8 | 1.0 |
| 3299 | 1.413731 | 1.0 | 3.766832 | 4.183655 | 8 | 1.0 |
3300 rows × 6 columns
Step 2: Model Setup & Priors¶
In [24]:
Copied!
model_reg_v_ex3_A1 = hssm.HSSM(
data=combined_dataset2,
include=[
{
"name": "v",
"formula": (
"v ~ 1 + (zSTN + zGPe)*sevScore + (1 + zSTN + zGPe | participant_id)"
),
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zSTN:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
model_reg_v_ex3_A1 = hssm.HSSM(
data=combined_dataset2,
include=[
{
"name": "v",
"formula": (
"v ~ 1 + (zSTN + zGPe)*sevScore + (1 + zSTN + zGPe | participant_id)"
),
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zSTN:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1.0},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
Model initialized successfully.
In [25]:
Copied!
samples_model_reg_v_ex3_A1 = model_reg_v_ex3_A1.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
samples_model_reg_v_ex3_A1 = model_reg_v_ex3_A1.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
Using default initvals.
NUTS[numpyro]: [a, z, t, v_Intercept, v_zSTN, v_zGPe, v_sevScore, v_zSTN:sevScore, v_zGPe:sevScore, v_1|participant_id_sigma, v_1|participant_id_offset, v_zSTN|participant_id_sigma, v_zSTN|participant_id_offset, v_zGPe|participant_id_sigma, v_zGPe|participant_id_offset]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:06<02:09, 7.32it/s, 511 steps of size 5.65e-03. acc. prob=0.74]
warmup: 5%|▌ | 50/1000 [00:21<02:09, 7.32it/s, 1023 steps of size 4.32e-03. acc. prob=0.75]
warmup: 9%|▉ | 91/1000 [00:21<03:51, 3.93it/s, 1023 steps of size 6.49e-03. acc. prob=0.76]
warmup: 9%|▉ | 92/1000 [00:21<03:51, 3.92it/s, 511 steps of size 7.99e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:21<03:52, 3.91it/s, 511 steps of size 4.18e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:22<04:02, 3.74it/s, 1023 steps of size 5.41e-03. acc. prob=0.76]
warmup: 10%|▉ | 95/1000 [00:22<04:03, 3.72it/s, 511 steps of size 5.35e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:23<04:20, 3.48it/s, 1023 steps of size 7.99e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:23<04:19, 3.48it/s, 511 steps of size 1.09e-02. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:24<04:21, 3.45it/s, 1023 steps of size 5.13e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:24<04:52, 3.07it/s, 1023 steps of size 6.79e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:25<04:47, 3.13it/s, 511 steps of size 9.76e-02. acc. prob=0.76]
warmup: 10%|█ | 105/1000 [00:25<02:46, 5.39it/s, 63 steps of size 9.37e-02. acc. prob=0.76]
warmup: 11%|█ | 111/1000 [00:25<01:30, 9.81it/s, 15 steps of size 2.80e-02. acc. prob=0.76]
warmup: 11%|█▏ | 113/1000 [00:25<01:25, 10.34it/s, 127 steps of size 8.05e-02. acc. prob=0.76]
warmup: 12%|█▏ | 118/1000 [00:25<00:58, 15.01it/s, 31 steps of size 1.63e-01. acc. prob=0.77]
warmup: 12%|█▏ | 123/1000 [00:25<00:46, 18.71it/s, 127 steps of size 8.31e-02. acc. prob=0.77]
warmup: 13%|█▎ | 128/1000 [00:25<00:38, 22.71it/s, 63 steps of size 1.26e-01. acc. prob=0.77]
warmup: 13%|█▎ | 132/1000 [00:26<00:35, 24.37it/s, 127 steps of size 6.75e-02. acc. prob=0.77]
warmup: 14%|█▎ | 136/1000 [00:26<00:32, 26.96it/s, 31 steps of size 3.62e-01. acc. prob=0.78]
warmup: 14%|█▍ | 139/1000 [00:26<00:32, 26.58it/s, 63 steps of size 1.24e-01. acc. prob=0.77]
warmup: 14%|█▍ | 144/1000 [00:26<00:27, 31.19it/s, 63 steps of size 9.56e-02. acc. prob=0.77]
warmup: 15%|█▍ | 148/1000 [00:26<00:26, 31.91it/s, 63 steps of size 1.16e-01. acc. prob=0.77]
warmup: 15%|█▌ | 150/1000 [00:26<00:33, 25.07it/s, 255 steps of size 4.87e-02. acc. prob=0.77]
warmup: 15%|█▌ | 154/1000 [00:26<00:30, 27.93it/s, 31 steps of size 1.68e-01. acc. prob=0.77]
warmup: 16%|█▌ | 156/1000 [00:26<00:37, 22.52it/s, 255 steps of size 3.75e-02. acc. prob=0.77]
warmup: 16%|█▌ | 158/1000 [00:27<00:39, 21.43it/s, 63 steps of size 1.21e-01. acc. prob=0.77]
warmup: 16%|█▋ | 164/1000 [00:27<00:34, 24.56it/s, 255 steps of size 4.33e-02. acc. prob=0.77]
warmup: 17%|█▋ | 166/1000 [00:27<00:35, 23.20it/s, 63 steps of size 1.49e-01. acc. prob=0.77]
warmup: 17%|█▋ | 172/1000 [00:27<00:26, 31.28it/s, 31 steps of size 2.25e-01. acc. prob=0.77]
warmup: 18%|█▊ | 177/1000 [00:27<00:24, 33.16it/s, 63 steps of size 1.25e-01. acc. prob=0.77]
warmup: 18%|█▊ | 184/1000 [00:27<00:22, 36.06it/s, 127 steps of size 1.38e-01. acc. prob=0.77]
warmup: 19%|█▉ | 190/1000 [00:27<00:20, 40.07it/s, 31 steps of size 1.28e-01. acc. prob=0.78]
warmup: 19%|█▉ | 193/1000 [00:27<00:21, 37.04it/s, 127 steps of size 7.95e-02. acc. prob=0.77]
warmup: 20%|█▉ | 197/1000 [00:28<00:21, 37.03it/s, 31 steps of size 2.35e-01. acc. prob=0.78]
warmup: 20%|██ | 202/1000 [00:28<00:20, 39.28it/s, 31 steps of size 1.70e-01. acc. prob=0.78]
warmup: 21%|██ | 208/1000 [00:28<00:18, 43.87it/s, 31 steps of size 2.75e-01. acc. prob=0.78]
warmup: 21%|██▏ | 213/1000 [00:28<00:18, 43.69it/s, 63 steps of size 7.34e-02. acc. prob=0.78]
warmup: 22%|██▏ | 217/1000 [00:28<00:19, 41.03it/s, 63 steps of size 1.14e-01. acc. prob=0.78]
warmup: 22%|██▏ | 221/1000 [00:28<00:19, 40.12it/s, 31 steps of size 1.82e-01. acc. prob=0.78]
warmup: 23%|██▎ | 228/1000 [00:28<00:16, 46.06it/s, 31 steps of size 1.87e-01. acc. prob=0.78]
warmup: 23%|██▎ | 234/1000 [00:28<00:15, 48.16it/s, 63 steps of size 1.68e-01. acc. prob=0.78]
warmup: 24%|██▍ | 239/1000 [00:28<00:15, 48.58it/s, 15 steps of size 7.89e-02. acc. prob=0.78]
warmup: 24%|██▍ | 244/1000 [00:29<00:15, 48.11it/s, 31 steps of size 2.04e-01. acc. prob=0.78]
warmup: 25%|██▌ | 251/1000 [00:29<00:14, 53.10it/s, 15 steps of size 1.54e+00. acc. prob=0.78]
warmup: 26%|██▌ | 259/1000 [00:29<00:14, 51.59it/s, 127 steps of size 6.43e-02. acc. prob=0.78]
warmup: 26%|██▋ | 264/1000 [00:29<00:15, 49.04it/s, 31 steps of size 2.62e-01. acc. prob=0.78]
warmup: 27%|██▋ | 267/1000 [00:29<00:17, 41.82it/s, 63 steps of size 1.09e-01. acc. prob=0.78]
warmup: 27%|██▋ | 273/1000 [00:29<00:16, 43.58it/s, 127 steps of size 9.10e-02. acc. prob=0.78]
warmup: 28%|██▊ | 280/1000 [00:29<00:14, 49.78it/s, 15 steps of size 1.71e-01. acc. prob=0.78]
warmup: 29%|██▉ | 289/1000 [00:29<00:12, 58.59it/s, 31 steps of size 3.29e-01. acc. prob=0.78]
warmup: 29%|██▉ | 294/1000 [00:30<00:13, 50.76it/s, 127 steps of size 6.35e-02. acc. prob=0.78]
warmup: 30%|██▉ | 298/1000 [00:30<00:14, 47.92it/s, 15 steps of size 2.67e-01. acc. prob=0.78]
warmup: 30%|███ | 305/1000 [00:30<00:14, 49.18it/s, 63 steps of size 1.56e-01. acc. prob=0.78]
warmup: 31%|███ | 312/1000 [00:30<00:13, 50.10it/s, 63 steps of size 1.47e-01. acc. prob=0.78]
warmup: 32%|███▏ | 319/1000 [00:30<00:12, 55.11it/s, 31 steps of size 1.17e-01. acc. prob=0.78]
warmup: 32%|███▏ | 323/1000 [00:30<00:15, 42.84it/s, 127 steps of size 8.48e-02. acc. prob=0.78]
warmup: 33%|███▎ | 327/1000 [00:30<00:16, 41.04it/s, 63 steps of size 9.52e-02. acc. prob=0.78]
warmup: 33%|███▎ | 332/1000 [00:30<00:16, 40.02it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 34%|███▍ | 338/1000 [00:31<00:15, 44.11it/s, 15 steps of size 2.26e-01. acc. prob=0.78]
warmup: 34%|███▍ | 344/1000 [00:31<00:14, 46.21it/s, 31 steps of size 2.22e-01. acc. prob=0.78]
warmup: 35%|███▍ | 349/1000 [00:31<00:14, 45.21it/s, 31 steps of size 1.55e-01. acc. prob=0.78]
warmup: 35%|███▌ | 354/1000 [00:31<00:14, 45.99it/s, 31 steps of size 1.72e-01. acc. prob=0.78]
warmup: 36%|███▋ | 363/1000 [00:31<00:11, 56.66it/s, 31 steps of size 1.72e-01. acc. prob=0.78]
warmup: 37%|███▋ | 370/1000 [00:31<00:10, 60.31it/s, 15 steps of size 2.67e-01. acc. prob=0.78]
warmup: 38%|███▊ | 378/1000 [00:31<00:09, 62.70it/s, 31 steps of size 1.23e-01. acc. prob=0.78]
warmup: 38%|███▊ | 384/1000 [00:31<00:10, 60.19it/s, 31 steps of size 1.36e-01. acc. prob=0.78]
warmup: 39%|███▉ | 390/1000 [00:31<00:10, 58.57it/s, 31 steps of size 2.26e-01. acc. prob=0.78]
warmup: 40%|███▉ | 398/1000 [00:32<00:09, 61.56it/s, 31 steps of size 1.15e-01. acc. prob=0.78]
warmup: 40%|████ | 405/1000 [00:32<00:09, 62.87it/s, 31 steps of size 2.21e-01. acc. prob=0.78]
warmup: 41%|████ | 412/1000 [00:32<00:09, 64.32it/s, 15 steps of size 2.37e-01. acc. prob=0.79]
warmup: 42%|████▏ | 419/1000 [00:32<00:09, 64.23it/s, 31 steps of size 2.10e-01. acc. prob=0.79]
warmup: 42%|████▎ | 425/1000 [00:32<00:09, 61.19it/s, 31 steps of size 1.95e-01. acc. prob=0.79]
warmup: 43%|████▎ | 434/1000 [00:32<00:08, 69.26it/s, 15 steps of size 1.47e-01. acc. prob=0.78]
warmup: 44%|████▍ | 441/1000 [00:32<00:08, 69.41it/s, 31 steps of size 1.58e-01. acc. prob=0.79]
warmup: 45%|████▍ | 448/1000 [00:32<00:08, 67.75it/s, 31 steps of size 1.43e-01. acc. prob=0.79]
warmup: 46%|████▌ | 458/1000 [00:32<00:08, 65.14it/s, 127 steps of size 6.26e-02. acc. prob=0.78]
warmup: 46%|████▌ | 462/1000 [00:33<00:09, 57.77it/s, 63 steps of size 9.48e-02. acc. prob=0.78]
warmup: 46%|████▋ | 465/1000 [00:33<00:13, 38.56it/s, 255 steps of size 3.50e-02. acc. prob=0.78]
warmup: 47%|████▋ | 467/1000 [00:33<00:15, 33.75it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 47%|████▋ | 471/1000 [00:33<00:15, 34.61it/s, 63 steps of size 1.64e-01. acc. prob=0.78]
warmup: 48%|████▊ | 478/1000 [00:33<00:12, 42.96it/s, 15 steps of size 6.15e-02. acc. prob=0.78]
warmup: 48%|████▊ | 483/1000 [00:33<00:11, 43.88it/s, 31 steps of size 2.27e-01. acc. prob=0.79]
warmup: 49%|████▉ | 491/1000 [00:33<00:09, 53.22it/s, 15 steps of size 2.64e-01. acc. prob=0.79]
warmup: 50%|████▉ | 498/1000 [00:33<00:09, 55.49it/s, 31 steps of size 7.30e-02. acc. prob=0.78]
sample: 50%|█████ | 504/1000 [00:34<00:09, 54.27it/s, 31 steps of size 1.53e-01. acc. prob=0.96]
sample: 51%|█████ | 510/1000 [00:34<00:08, 54.94it/s, 31 steps of size 1.53e-01. acc. prob=0.96]
sample: 52%|█████▏ | 516/1000 [00:34<00:08, 55.27it/s, 31 steps of size 1.53e-01. acc. prob=0.94]
sample: 52%|█████▏ | 523/1000 [00:34<00:08, 57.96it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 53%|█████▎ | 529/1000 [00:34<00:08, 57.50it/s, 31 steps of size 1.53e-01. acc. prob=0.81]
sample: 53%|█████▎ | 534/1000 [00:34<00:08, 54.16it/s, 31 steps of size 1.53e-01. acc. prob=0.81]
sample: 54%|█████▍ | 540/1000 [00:34<00:08, 54.68it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 55%|█████▍ | 547/1000 [00:34<00:08, 56.14it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 55%|█████▌ | 554/1000 [00:34<00:07, 57.38it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 56%|█████▌ | 561/1000 [00:35<00:07, 58.32it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 57%|█████▋ | 567/1000 [00:35<00:07, 57.75it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 57%|█████▋ | 572/1000 [00:35<00:07, 54.92it/s, 31 steps of size 1.53e-01. acc. prob=0.81]
sample: 58%|█████▊ | 578/1000 [00:35<00:07, 55.19it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 58%|█████▊ | 584/1000 [00:35<00:07, 55.39it/s, 31 steps of size 1.53e-01. acc. prob=0.79]
sample: 59%|█████▉ | 590/1000 [00:35<00:07, 55.33it/s, 31 steps of size 1.53e-01. acc. prob=0.80]
sample: 60%|█████▉ | 597/1000 [00:35<00:06, 58.19it/s, 15 steps of size 1.53e-01. acc. prob=0.81]
sample: 60%|██████ | 603/1000 [00:35<00:06, 57.23it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 61%|██████ | 609/1000 [00:35<00:06, 56.84it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 62%|██████▏ | 615/1000 [00:36<00:06, 56.42it/s, 31 steps of size 1.53e-01. acc. prob=0.81]
sample: 62%|██████▏ | 621/1000 [00:36<00:06, 56.10it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 63%|██████▎ | 627/1000 [00:36<00:06, 55.42it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 63%|██████▎ | 634/1000 [00:36<00:06, 57.08it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 64%|██████▍ | 640/1000 [00:36<00:06, 56.46it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 65%|██████▍ | 646/1000 [00:36<00:06, 55.23it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 65%|██████▌ | 653/1000 [00:36<00:06, 57.43it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 66%|██████▌ | 659/1000 [00:36<00:06, 56.75it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 66%|██████▋ | 665/1000 [00:36<00:05, 56.58it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 67%|██████▋ | 671/1000 [00:37<00:05, 56.98it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 68%|██████▊ | 678/1000 [00:37<00:05, 59.52it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 68%|██████▊ | 684/1000 [00:37<00:05, 58.40it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 69%|██████▉ | 691/1000 [00:37<00:05, 58.92it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 70%|██████▉ | 697/1000 [00:37<00:05, 57.64it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 70%|███████ | 703/1000 [00:37<00:05, 57.38it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 71%|███████ | 709/1000 [00:37<00:05, 58.10it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 72%|███████▏ | 715/1000 [00:37<00:04, 57.31it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 72%|███████▏ | 721/1000 [00:37<00:04, 56.51it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 73%|███████▎ | 727/1000 [00:37<00:04, 56.40it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 73%|███████▎ | 733/1000 [00:38<00:04, 56.05it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 74%|███████▍ | 739/1000 [00:38<00:04, 56.32it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 74%|███████▍ | 745/1000 [00:38<00:04, 55.98it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 75%|███████▌ | 751/1000 [00:38<00:04, 56.08it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 76%|███████▌ | 757/1000 [00:38<00:04, 55.56it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 76%|███████▋ | 763/1000 [00:38<00:04, 55.91it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 77%|███████▋ | 770/1000 [00:38<00:04, 57.46it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 78%|███████▊ | 776/1000 [00:38<00:03, 57.17it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 78%|███████▊ | 782/1000 [00:38<00:03, 56.60it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 79%|███████▉ | 788/1000 [00:39<00:03, 56.00it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 79%|███████▉ | 794/1000 [00:39<00:03, 56.10it/s, 31 steps of size 1.53e-01. acc. prob=0.84]
sample: 80%|████████ | 800/1000 [00:39<00:03, 56.25it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 81%|████████ | 806/1000 [00:39<00:03, 55.66it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 81%|████████▏ | 813/1000 [00:39<00:03, 57.23it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 82%|████████▏ | 819/1000 [00:39<00:03, 56.85it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 82%|████████▎ | 825/1000 [00:39<00:03, 56.74it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 83%|████████▎ | 831/1000 [00:39<00:02, 56.55it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 84%|████████▎ | 837/1000 [00:39<00:02, 56.27it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 84%|████████▍ | 843/1000 [00:40<00:02, 56.03it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 85%|████████▍ | 849/1000 [00:40<00:02, 56.05it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 86%|████████▌ | 855/1000 [00:40<00:02, 56.99it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 86%|████████▌ | 861/1000 [00:40<00:02, 57.19it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 87%|████████▋ | 867/1000 [00:40<00:02, 55.48it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 87%|████████▋ | 873/1000 [00:40<00:02, 55.68it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 88%|████████▊ | 878/1000 [00:40<00:02, 53.75it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 88%|████████▊ | 883/1000 [00:40<00:02, 52.19it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 89%|████████▉ | 889/1000 [00:40<00:02, 53.50it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 90%|████████▉ | 895/1000 [00:41<00:01, 53.09it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 90%|█████████ | 901/1000 [00:41<00:01, 53.64it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 91%|█████████ | 907/1000 [00:41<00:01, 55.42it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 91%|█████████▏| 913/1000 [00:41<00:01, 55.21it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 92%|█████████▏| 919/1000 [00:41<00:01, 55.18it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 93%|█████████▎| 926/1000 [00:41<00:01, 56.96it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 93%|█████████▎| 932/1000 [00:41<00:01, 56.25it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 94%|█████████▍| 939/1000 [00:41<00:01, 57.82it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 95%|█████████▍| 946/1000 [00:41<00:00, 60.15it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 95%|█████████▌| 953/1000 [00:42<00:00, 59.94it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 96%|█████████▌| 959/1000 [00:42<00:00, 58.65it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 96%|█████████▋| 965/1000 [00:42<00:00, 58.91it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 97%|█████████▋| 971/1000 [00:42<00:00, 57.35it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 98%|█████████▊| 977/1000 [00:42<00:00, 57.94it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 98%|█████████▊| 983/1000 [00:42<00:00, 56.82it/s, 31 steps of size 1.53e-01. acc. prob=0.82]
sample: 99%|█████████▉| 989/1000 [00:42<00:00, 56.05it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 100%|█████████▉| 995/1000 [00:42<00:00, 56.19it/s, 31 steps of size 1.53e-01. acc. prob=0.83]
sample: 100%|██████████| 1000/1000 [00:42<00:00, 23.34it/s, 15 steps of size 1.53e-01. acc. prob=0.83]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:09<02:51, 5.53it/s, 255 steps of size 1.19e-02. acc. prob=0.74]
warmup: 10%|█ | 100/1000 [00:26<04:10, 3.59it/s, 511 steps of size 7.39e-03. acc. prob=0.76]
warmup: 15%|█▌ | 150/1000 [00:27<02:19, 6.08it/s, 31 steps of size 1.20e-01. acc. prob=0.77]
warmup: 20%|██ | 200/1000 [00:29<01:27, 9.18it/s, 31 steps of size 2.34e-01. acc. prob=0.78]
warmup: 25%|██▌ | 250/1000 [00:29<00:56, 13.23it/s, 31 steps of size 1.91e-01. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:30<00:40, 17.34it/s, 15 steps of size 2.05e-01. acc. prob=0.78]
warmup: 35%|███▌ | 350/1000 [00:31<00:28, 22.45it/s, 63 steps of size 8.13e-02. acc. prob=0.78]
warmup: 40%|████ | 400/1000 [00:33<00:23, 26.01it/s, 63 steps of size 9.38e-02. acc. prob=0.78]
warmup: 45%|████▌ | 450/1000 [00:34<00:17, 31.08it/s, 31 steps of size 1.89e-01. acc. prob=0.79]
warmup: 50%|█████ | 500/1000 [00:35<00:15, 32.70it/s, 15 steps of size 1.34e-01. acc. prob=0.79]
sample: 55%|█████▌ | 550/1000 [00:36<00:11, 37.52it/s, 31 steps of size 1.34e-01. acc. prob=0.92]
sample: 60%|██████ | 600/1000 [00:37<00:09, 41.78it/s, 31 steps of size 1.34e-01. acc. prob=0.93]
sample: 65%|██████▌ | 650/1000 [00:38<00:07, 44.87it/s, 31 steps of size 1.34e-01. acc. prob=0.92]
sample: 70%|███████ | 700/1000 [00:39<00:06, 47.66it/s, 31 steps of size 1.34e-01. acc. prob=0.90]
sample: 75%|███████▌ | 750/1000 [00:39<00:05, 49.68it/s, 31 steps of size 1.34e-01. acc. prob=0.91]
sample: 80%|████████ | 800/1000 [00:40<00:03, 51.43it/s, 31 steps of size 1.34e-01. acc. prob=0.90]
sample: 85%|████████▌ | 850/1000 [00:41<00:02, 52.62it/s, 31 steps of size 1.34e-01. acc. prob=0.90]
sample: 90%|█████████ | 900/1000 [00:42<00:01, 53.47it/s, 31 steps of size 1.34e-01. acc. prob=0.90]
sample: 95%|█████████▌| 950/1000 [00:43<00:00, 51.46it/s, 31 steps of size 1.34e-01. acc. prob=0.88]
sample: 100%|██████████| 1000/1000 [00:44<00:00, 52.42it/s, 31 steps of size 1.34e-01. acc. prob=0.88]
sample: 100%|██████████| 1000/1000 [00:44<00:00, 22.44it/s, 31 steps of size 1.34e-01. acc. prob=0.88]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:10<03:18, 4.80it/s, 1023 steps of size 2.91e-03. acc. prob=0.73]
warmup: 5%|▌ | 50/1000 [00:23<03:18, 4.80it/s, 255 steps of size 2.54e-03. acc. prob=0.75]
warmup: 8%|▊ | 84/1000 [00:23<04:32, 3.36it/s, 1023 steps of size 3.67e-03. acc. prob=0.75]
warmup: 8%|▊ | 85/1000 [00:24<04:37, 3.30it/s, 1023 steps of size 5.61e-03. acc. prob=0.75]
warmup: 9%|▊ | 86/1000 [00:24<04:36, 3.30it/s, 511 steps of size 7.67e-03. acc. prob=0.76]
warmup: 9%|▊ | 87/1000 [00:24<04:31, 3.37it/s, 255 steps of size 5.26e-03. acc. prob=0.75]
warmup: 9%|▉ | 88/1000 [00:25<04:44, 3.20it/s, 1023 steps of size 7.00e-03. acc. prob=0.76]
warmup: 9%|▉ | 89/1000 [00:25<04:43, 3.22it/s, 511 steps of size 1.02e-02. acc. prob=0.76]
warmup: 9%|▉ | 90/1000 [00:25<04:25, 3.42it/s, 184 steps of size 2.35e-03. acc. prob=0.75]
warmup: 9%|▉ | 91/1000 [00:26<04:54, 3.09it/s, 1023 steps of size 3.48e-03. acc. prob=0.75]
warmup: 9%|▉ | 92/1000 [00:26<05:25, 2.79it/s, 1023 steps of size 5.24e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:27<05:14, 2.88it/s, 511 steps of size 6.24e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:27<05:02, 2.99it/s, 511 steps of size 7.11e-03. acc. prob=0.76]
warmup: 10%|▉ | 95/1000 [00:27<04:52, 3.09it/s, 511 steps of size 2.99e-03. acc. prob=0.75]
warmup: 10%|▉ | 96/1000 [00:28<05:47, 2.60it/s, 1023 steps of size 4.48e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:28<06:42, 2.24it/s, 1023 steps of size 5.89e-03. acc. prob=0.76]
warmup: 10%|▉ | 98/1000 [00:29<06:04, 2.47it/s, 511 steps of size 5.32e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:29<05:35, 2.68it/s, 511 steps of size 7.58e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:29<05:14, 2.87it/s, 511 steps of size 2.80e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:30<06:12, 2.41it/s, 1023 steps of size 4.03e-02. acc. prob=0.76]
warmup: 10%|█ | 103/1000 [00:30<03:46, 3.96it/s, 63 steps of size 1.36e-01. acc. prob=0.76]
warmup: 11%|█ | 109/1000 [00:30<01:28, 10.09it/s, 63 steps of size 2.22e-01. acc. prob=0.76]
warmup: 12%|█▏ | 116/1000 [00:30<00:52, 16.93it/s, 127 steps of size 8.85e-02. acc. prob=0.76]
warmup: 12%|█▏ | 122/1000 [00:30<00:40, 21.66it/s, 127 steps of size 7.82e-02. acc. prob=0.77]
warmup: 13%|█▎ | 127/1000 [00:31<00:35, 24.46it/s, 127 steps of size 5.78e-02. acc. prob=0.77]
warmup: 13%|█▎ | 132/1000 [00:31<00:30, 28.69it/s, 31 steps of size 9.83e-02. acc. prob=0.77]
warmup: 14%|█▍ | 138/1000 [00:31<00:25, 34.33it/s, 31 steps of size 1.79e-01. acc. prob=0.77]
warmup: 15%|█▍ | 146/1000 [00:31<00:19, 43.71it/s, 15 steps of size 2.60e-01. acc. prob=0.78]
warmup: 15%|█▌ | 154/1000 [00:31<00:16, 49.89it/s, 63 steps of size 8.68e-02. acc. prob=0.77]
warmup: 16%|█▌ | 161/1000 [00:31<00:16, 50.21it/s, 127 steps of size 6.80e-02. acc. prob=0.77]
warmup: 16%|█▋ | 165/1000 [00:31<00:19, 43.30it/s, 127 steps of size 7.63e-02. acc. prob=0.77]
warmup: 17%|█▋ | 170/1000 [00:31<00:18, 43.94it/s, 31 steps of size 2.60e-01. acc. prob=0.77]
warmup: 18%|█▊ | 175/1000 [00:31<00:18, 44.02it/s, 31 steps of size 2.52e-01. acc. prob=0.77]
warmup: 18%|█▊ | 181/1000 [00:32<00:17, 47.30it/s, 15 steps of size 9.37e-02. acc. prob=0.77]
warmup: 18%|█▊ | 185/1000 [00:32<00:18, 43.55it/s, 127 steps of size 8.84e-02. acc. prob=0.77]
warmup: 19%|█▉ | 188/1000 [00:32<00:21, 37.11it/s, 127 steps of size 8.35e-02. acc. prob=0.77]
warmup: 19%|█▉ | 190/1000 [00:32<00:25, 32.02it/s, 127 steps of size 4.20e-02. acc. prob=0.77]
warmup: 19%|█▉ | 192/1000 [00:32<00:28, 28.13it/s, 63 steps of size 8.00e-02. acc. prob=0.77]
warmup: 20%|█▉ | 195/1000 [00:32<00:29, 27.11it/s, 127 steps of size 5.86e-02. acc. prob=0.77]
warmup: 20%|██ | 201/1000 [00:32<00:23, 33.65it/s, 63 steps of size 8.07e-02. acc. prob=0.77]
warmup: 21%|██ | 207/1000 [00:32<00:20, 38.94it/s, 31 steps of size 1.56e-01. acc. prob=0.78]
warmup: 21%|██ | 212/1000 [00:32<00:18, 41.81it/s, 31 steps of size 1.97e-01. acc. prob=0.78]
warmup: 22%|██▏ | 221/1000 [00:33<00:15, 51.37it/s, 63 steps of size 9.96e-02. acc. prob=0.78]
warmup: 23%|██▎ | 227/1000 [00:33<00:14, 52.62it/s, 31 steps of size 2.14e-01. acc. prob=0.78]
warmup: 24%|██▎ | 235/1000 [00:33<00:13, 57.96it/s, 31 steps of size 7.60e-02. acc. prob=0.78]
warmup: 24%|██▍ | 242/1000 [00:33<00:12, 58.80it/s, 31 steps of size 1.68e-01. acc. prob=0.78]
warmup: 25%|██▍ | 249/1000 [00:33<00:12, 59.32it/s, 31 steps of size 1.80e-01. acc. prob=0.78]
warmup: 26%|██▌ | 255/1000 [00:33<00:12, 57.68it/s, 31 steps of size 1.82e-01. acc. prob=0.78]
warmup: 26%|██▌ | 258/1000 [00:33<00:18, 40.73it/s, 255 steps of size 4.35e-02. acc. prob=0.78]
warmup: 26%|██▌ | 260/1000 [00:33<00:21, 34.01it/s, 63 steps of size 1.21e-01. acc. prob=0.78]
warmup: 26%|██▋ | 263/1000 [00:34<00:24, 30.39it/s, 127 steps of size 4.85e-02. acc. prob=0.78]
warmup: 26%|██▋ | 265/1000 [00:34<00:27, 27.13it/s, 63 steps of size 1.43e-01. acc. prob=0.78]
warmup: 27%|██▋ | 271/1000 [00:34<00:23, 31.26it/s, 127 steps of size 6.26e-02. acc. prob=0.78]
warmup: 28%|██▊ | 276/1000 [00:34<00:20, 35.53it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 28%|██▊ | 280/1000 [00:34<00:20, 35.80it/s, 31 steps of size 1.77e-01. acc. prob=0.78]
warmup: 28%|██▊ | 284/1000 [00:34<00:20, 34.12it/s, 63 steps of size 9.87e-02. acc. prob=0.78]
warmup: 29%|██▉ | 289/1000 [00:34<00:19, 36.34it/s, 63 steps of size 1.24e-01. acc. prob=0.78]
warmup: 29%|██▉ | 294/1000 [00:34<00:18, 38.90it/s, 31 steps of size 1.06e-01. acc. prob=0.78]
warmup: 30%|██▉ | 299/1000 [00:35<00:19, 35.11it/s, 127 steps of size 6.40e-02. acc. prob=0.78]
warmup: 30%|███ | 301/1000 [00:35<00:22, 30.66it/s, 127 steps of size 1.03e-01. acc. prob=0.78]
warmup: 31%|███ | 306/1000 [00:35<00:19, 34.74it/s, 31 steps of size 1.49e-01. acc. prob=0.78]
warmup: 31%|███ | 310/1000 [00:35<00:19, 35.36it/s, 31 steps of size 2.12e-01. acc. prob=0.78]
warmup: 31%|███▏ | 314/1000 [00:35<00:18, 36.51it/s, 63 steps of size 1.09e-01. acc. prob=0.78]
warmup: 32%|███▏ | 319/1000 [00:35<00:17, 39.13it/s, 31 steps of size 1.27e-01. acc. prob=0.78]
warmup: 32%|███▏ | 324/1000 [00:35<00:16, 40.16it/s, 31 steps of size 1.57e-01. acc. prob=0.78]
warmup: 33%|███▎ | 329/1000 [00:35<00:16, 41.81it/s, 31 steps of size 1.28e-01. acc. prob=0.78]
warmup: 33%|███▎ | 333/1000 [00:35<00:16, 40.16it/s, 63 steps of size 1.65e-01. acc. prob=0.78]
warmup: 34%|███▍ | 339/1000 [00:36<00:15, 43.66it/s, 63 steps of size 1.14e-01. acc. prob=0.78]
warmup: 34%|███▍ | 344/1000 [00:36<00:15, 43.21it/s, 63 steps of size 1.28e-01. acc. prob=0.78]
warmup: 35%|███▍ | 349/1000 [00:36<00:15, 43.26it/s, 63 steps of size 1.04e-01. acc. prob=0.78]
warmup: 35%|███▌ | 353/1000 [00:36<00:16, 40.30it/s, 127 steps of size 6.97e-02. acc. prob=0.78]
warmup: 36%|███▌ | 358/1000 [00:36<00:16, 40.11it/s, 63 steps of size 8.02e-02. acc. prob=0.78]
warmup: 36%|███▋ | 363/1000 [00:36<00:15, 41.94it/s, 31 steps of size 1.69e-01. acc. prob=0.78]
warmup: 37%|███▋ | 369/1000 [00:36<00:14, 44.86it/s, 31 steps of size 1.37e-01. acc. prob=0.78]
warmup: 38%|███▊ | 376/1000 [00:36<00:12, 50.36it/s, 31 steps of size 8.22e-02. acc. prob=0.78]
warmup: 38%|███▊ | 380/1000 [00:37<00:13, 46.67it/s, 31 steps of size 1.68e-01. acc. prob=0.78]
warmup: 38%|███▊ | 385/1000 [00:37<00:13, 45.29it/s, 31 steps of size 1.48e-01. acc. prob=0.78]
warmup: 39%|███▉ | 390/1000 [00:37<00:13, 45.37it/s, 31 steps of size 1.77e-01. acc. prob=0.78]
warmup: 40%|███▉ | 396/1000 [00:37<00:13, 45.06it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 40%|████ | 402/1000 [00:37<00:12, 46.94it/s, 31 steps of size 1.29e-01. acc. prob=0.78]
warmup: 41%|████ | 406/1000 [00:37<00:13, 43.93it/s, 63 steps of size 9.92e-02. acc. prob=0.78]
warmup: 41%|████ | 411/1000 [00:37<00:13, 44.60it/s, 63 steps of size 9.75e-02. acc. prob=0.78]
warmup: 42%|████▏ | 416/1000 [00:37<00:12, 45.28it/s, 31 steps of size 1.00e-01. acc. prob=0.78]
warmup: 42%|████▏ | 422/1000 [00:37<00:12, 45.91it/s, 63 steps of size 9.42e-02. acc. prob=0.78]
warmup: 43%|████▎ | 426/1000 [00:38<00:13, 43.91it/s, 31 steps of size 1.14e-01. acc. prob=0.78]
warmup: 43%|████▎ | 430/1000 [00:38<00:13, 42.82it/s, 63 steps of size 1.54e-01. acc. prob=0.78]
warmup: 43%|████▎ | 433/1000 [00:38<00:16, 35.17it/s, 63 steps of size 8.67e-02. acc. prob=0.78]
warmup: 44%|████▍ | 438/1000 [00:38<00:14, 37.63it/s, 31 steps of size 1.73e-01. acc. prob=0.78]
warmup: 44%|████▍ | 444/1000 [00:38<00:13, 41.67it/s, 63 steps of size 1.01e-01. acc. prob=0.78]
warmup: 45%|████▍ | 449/1000 [00:38<00:13, 40.94it/s, 63 steps of size 6.86e-02. acc. prob=0.78]
warmup: 45%|████▌ | 454/1000 [00:38<00:12, 43.20it/s, 31 steps of size 1.49e-01. acc. prob=0.78]
warmup: 46%|████▌ | 459/1000 [00:38<00:13, 41.12it/s, 127 steps of size 6.34e-02. acc. prob=0.78]
warmup: 46%|████▋ | 463/1000 [00:39<00:15, 35.04it/s, 127 steps of size 8.27e-02. acc. prob=0.78]
warmup: 47%|████▋ | 468/1000 [00:39<00:13, 38.00it/s, 15 steps of size 1.44e-01. acc. prob=0.78]
warmup: 47%|████▋ | 471/1000 [00:39<00:17, 30.57it/s, 255 steps of size 5.18e-02. acc. prob=0.78]
warmup: 48%|████▊ | 475/1000 [00:39<00:16, 32.53it/s, 15 steps of size 2.86e-01. acc. prob=0.78]
warmup: 48%|████▊ | 482/1000 [00:39<00:13, 38.58it/s, 127 steps of size 8.48e-02. acc. prob=0.78]
warmup: 49%|████▉ | 488/1000 [00:39<00:12, 42.11it/s, 31 steps of size 2.71e-02. acc. prob=0.78]
warmup: 49%|████▉ | 489/1000 [00:39<00:16, 30.87it/s, 255 steps of size 4.63e-02. acc. prob=0.78]
warmup: 49%|████▉ | 491/1000 [00:39<00:18, 27.66it/s, 63 steps of size 1.25e-01. acc. prob=0.78]
warmup: 50%|████▉ | 499/1000 [00:40<00:12, 38.72it/s, 63 steps of size 1.04e-01. acc. prob=0.78]
sample: 51%|█████ | 506/1000 [00:40<00:10, 44.95it/s, 31 steps of size 1.27e-01. acc. prob=0.91]
sample: 51%|█████ | 512/1000 [00:40<00:10, 48.29it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 52%|█████▏ | 518/1000 [00:40<00:09, 50.96it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 52%|█████▏ | 523/1000 [00:40<00:09, 49.76it/s, 31 steps of size 1.27e-01. acc. prob=0.95]
sample: 53%|█████▎ | 529/1000 [00:40<00:09, 51.60it/s, 31 steps of size 1.27e-01. acc. prob=0.95]
sample: 54%|█████▎ | 535/1000 [00:40<00:08, 52.42it/s, 31 steps of size 1.27e-01. acc. prob=0.95]
sample: 54%|█████▍ | 541/1000 [00:40<00:08, 53.44it/s, 31 steps of size 1.27e-01. acc. prob=0.96]
sample: 55%|█████▍ | 547/1000 [00:40<00:08, 53.96it/s, 31 steps of size 1.27e-01. acc. prob=0.95]
sample: 55%|█████▌ | 553/1000 [00:41<00:08, 54.53it/s, 31 steps of size 1.27e-01. acc. prob=0.95]
sample: 56%|█████▌ | 559/1000 [00:41<00:08, 55.08it/s, 31 steps of size 1.27e-01. acc. prob=0.95]
sample: 56%|█████▋ | 565/1000 [00:41<00:07, 56.05it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 57%|█████▋ | 571/1000 [00:41<00:07, 56.16it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 58%|█████▊ | 577/1000 [00:41<00:07, 56.18it/s, 31 steps of size 1.27e-01. acc. prob=0.92]
sample: 58%|█████▊ | 583/1000 [00:41<00:07, 56.35it/s, 31 steps of size 1.27e-01. acc. prob=0.92]
sample: 59%|█████▉ | 588/1000 [00:41<00:07, 53.33it/s, 63 steps of size 1.27e-01. acc. prob=0.92]
sample: 59%|█████▉ | 593/1000 [00:41<00:07, 51.32it/s, 31 steps of size 1.27e-01. acc. prob=0.92]
sample: 60%|█████▉ | 599/1000 [00:41<00:07, 52.68it/s, 31 steps of size 1.27e-01. acc. prob=0.92]
sample: 60%|██████ | 605/1000 [00:41<00:07, 53.69it/s, 31 steps of size 1.27e-01. acc. prob=0.92]
sample: 61%|██████ | 611/1000 [00:42<00:07, 54.21it/s, 31 steps of size 1.27e-01. acc. prob=0.92]
sample: 62%|██████▏ | 617/1000 [00:42<00:06, 54.74it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 62%|██████▏ | 623/1000 [00:42<00:06, 55.65it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 63%|██████▎ | 629/1000 [00:42<00:06, 56.11it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 64%|██████▎ | 635/1000 [00:42<00:06, 56.21it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 64%|██████▍ | 641/1000 [00:42<00:06, 56.03it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 65%|██████▍ | 647/1000 [00:42<00:06, 55.81it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 65%|██████▌ | 653/1000 [00:42<00:06, 55.71it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 66%|██████▌ | 659/1000 [00:42<00:06, 55.31it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 66%|██████▋ | 665/1000 [00:43<00:06, 55.07it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 67%|██████▋ | 671/1000 [00:43<00:06, 54.70it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 68%|██████▊ | 677/1000 [00:43<00:05, 55.00it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 68%|██████▊ | 682/1000 [00:43<00:05, 53.57it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 69%|██████▊ | 687/1000 [00:43<00:06, 51.42it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 69%|██████▉ | 693/1000 [00:43<00:05, 51.37it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 70%|██████▉ | 699/1000 [00:43<00:05, 51.51it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 70%|███████ | 705/1000 [00:43<00:05, 52.10it/s, 31 steps of size 1.27e-01. acc. prob=0.94]
sample: 71%|███████ | 709/1000 [00:43<00:06, 47.69it/s, 31 steps of size 1.27e-01. acc. prob=0.93]
sample: 72%|███████▏ | 721/1000 [00:44<00:04, 66.69it/s, 17 steps of size 1.27e-01. acc. prob=0.89]
sample: 74%|███████▍ | 738/1000 [00:44<00:02, 95.31it/s, 8 steps of size 1.27e-01. acc. prob=0.83]
sample: 75%|███████▍ | 748/1000 [00:44<00:02, 96.12it/s, 3 steps of size 1.27e-01. acc. prob=0.80]
sample: 76%|███████▋ | 765/1000 [00:44<00:02, 117.39it/s, 9 steps of size 1.27e-01. acc. prob=0.75]
sample: 80%|████████ | 800/1000 [00:44<00:01, 184.65it/s, 9 steps of size 1.27e-01. acc. prob=0.67]
sample: 83%|████████▎ | 831/1000 [00:44<00:00, 220.93it/s, 7 steps of size 1.27e-01. acc. prob=0.61]
sample: 85%|████████▍ | 849/1000 [00:44<00:00, 204.58it/s, 15 steps of size 1.27e-01. acc. prob=0.59]
sample: 86%|████████▋ | 863/1000 [00:44<00:00, 185.45it/s, 5 steps of size 1.27e-01. acc. prob=0.57]
sample: 88%|████████▊ | 877/1000 [00:44<00:00, 164.06it/s, 31 steps of size 1.27e-01. acc. prob=0.55]
sample: 89%|████████▉ | 889/1000 [00:44<00:00, 150.22it/s, 31 steps of size 1.27e-01. acc. prob=0.54]
sample: 90%|████████▉ | 896/1000 [00:45<00:00, 123.25it/s, 31 steps of size 1.27e-01. acc. prob=0.53]
sample: 90%|█████████ | 905/1000 [00:45<00:00, 98.14it/s, 127 steps of size 1.27e-01. acc. prob=0.52]
sample: 91%|█████████ | 912/1000 [00:45<00:00, 90.79it/s, 7 steps of size 1.27e-01. acc. prob=0.51]
sample: 92%|█████████▏| 920/1000 [00:45<00:01, 79.92it/s, 127 steps of size 1.27e-01. acc. prob=0.50]
sample: 93%|█████████▎| 928/1000 [00:45<00:00, 76.58it/s, 31 steps of size 1.27e-01. acc. prob=0.50]
sample: 93%|█████████▎| 934/1000 [00:45<00:00, 70.99it/s, 31 steps of size 1.27e-01. acc. prob=0.49]
sample: 94%|█████████▍| 940/1000 [00:45<00:00, 65.95it/s, 31 steps of size 1.27e-01. acc. prob=0.49]
sample: 95%|█████████▍| 947/1000 [00:45<00:00, 64.57it/s, 31 steps of size 1.27e-01. acc. prob=0.49]
sample: 96%|█████████▌| 955/1000 [00:46<00:00, 68.56it/s, 15 steps of size 1.27e-01. acc. prob=0.48]
sample: 96%|█████████▌| 962/1000 [00:46<00:00, 63.11it/s, 63 steps of size 1.27e-01. acc. prob=0.48]
sample: 97%|█████████▋| 968/1000 [00:46<00:00, 61.06it/s, 31 steps of size 1.27e-01. acc. prob=0.49]
sample: 97%|█████████▋| 974/1000 [00:46<00:00, 59.49it/s, 31 steps of size 1.27e-01. acc. prob=0.50]
sample: 98%|█████████▊| 980/1000 [00:46<00:00, 58.26it/s, 31 steps of size 1.27e-01. acc. prob=0.50]
sample: 99%|█████████▊| 986/1000 [00:46<00:00, 57.47it/s, 31 steps of size 1.27e-01. acc. prob=0.51]
sample: 99%|█████████▉| 992/1000 [00:46<00:00, 56.85it/s, 31 steps of size 1.27e-01. acc. prob=0.51]
sample: 100%|█████████▉| 998/1000 [00:46<00:00, 56.45it/s, 31 steps of size 1.27e-01. acc. prob=0.52]
sample: 100%|██████████| 1000/1000 [00:46<00:00, 21.34it/s, 31 steps of size 1.27e-01. acc. prob=0.52]
There were 133 divergences after tuning. Increase `target_accept` or reparameterize.
We recommend running at least 4 chains for robust computation of convergence diagnostics
The rhat statistic is larger than 1.01 for some parameters. This indicates problems during sampling. See https://arxiv.org/abs/1903.08008 for details
The effective sample size per chain is smaller than 100 for some parameters. A higher number is needed for reliable rhat and ess computation. See https://arxiv.org/abs/1903.08008 for details
In [26]:
Copied!
#### posterior statistics of non-centered model
az.summary(samples_model_reg_v_ex3_A1)
#### posterior statistics of non-centered model
az.summary(samples_model_reg_v_ex3_A1)
Out[26]:
| mean | sd | eti89_lb | eti89_ub | ess_bulk | ess_tail | r_hat | mcse_mean | mcse_sd | |
|---|---|---|---|---|---|---|---|---|---|
| v_1|participant_id_sigma | 0.039 | 0.033 | 0.0034 | 0.099 | 253 | 449 | 1.01 | 0.002 | 0.003 |
| v_zGPe | 0.284 | 0.029 | 0.24 | 0.32 | 75 | 157 | 1.04 | 0.0027 | 0.003 |
| z | 0.49 | 0.0077 | 0.48 | 0.5 | 553 | 691 | 1.01 | 0.00032 | 0.00023 |
| v_zSTN|participant_id[1] | 0.02 | 0.05 | -0.023 | 0.13 | 8 | 5 | 1.27 | 0.022 | 0.019 |
| v_zSTN|participant_id[2] | 0.01 | 0.03 | -0.022 | 0.076 | 9 | 8 | 1.22 | 0.012 | 0.014 |
| v_zSTN|participant_id[3] | 0.01 | 0.02 | -0.022 | 0.054 | 16 | 18 | 1.14 | 0.0068 | 0.0063 |
| v_zSTN|participant_id[4] | -0.004 | 0.023 | -0.042 | 0.03 | 161 | 66 | 1.03 | 0.0026 | 0.0032 |
| v_zSTN|participant_id[5] | 0 | 0.03 | -0.047 | 0.04 | 17 | 8 | 1.13 | 0.007 | 0.0069 |
| v_zSTN|participant_id[6] | -0.02 | 0.04 | -0.12 | 0.016 | 8 | 7 | 1.26 | 0.017 | 0.017 |
| v_zSTN|participant_id[7] | -0.02 | 0.05 | -0.15 | 0.025 | 9 | 8 | 1.24 | 0.023 | 0.024 |
| v_zSTN|participant_id[8] | -0.05 | 0.09 | -0.27 | 0.0094 | 8 | 7 | 1.26 | 0.042 | 0.038 |
| v_sevScore | 0.042 | 0.086 | -0.07 | 0.19 | 326 | 457 | 1.01 | 0.0047 | 0.0046 |
| v_1|participant_id_offset[1] | -0.07 | 0.83 | -1.3 | 1.3 | 413 | 561 | 1.00 | 0.041 | 0.03 |
| v_1|participant_id_offset[2] | -0.05 | 0.86 | -1.4 | 1.3 | 318 | 221 | 1.01 | 0.048 | 0.034 |
| v_1|participant_id_offset[3] | 0.55 | 0.87 | -0.77 | 1.9 | 587 | 708 | 1.02 | 0.036 | 0.027 |
| v_1|participant_id_offset[4] | -0.26 | 0.81 | -1.5 | 1.1 | 323 | 789 | 1.01 | 0.044 | 0.034 |
| v_1|participant_id_offset[5] | 0.1 | 0.78 | -1.1 | 1.4 | 1487 | 798 | 1.03 | 0.019 | 0.016 |
| v_1|participant_id_offset[6] | -0.5 | 0.9 | -2 | 1.1 | 30 | 60 | 1.06 | 0.17 | 0.11 |
| v_1|participant_id_offset[7] | 0.15 | 0.81 | -1.1 | 1.4 | 821 | 714 | 1.03 | 0.027 | 0.021 |
| v_1|participant_id_offset[8] | -0.3 | 0.9 | -1.6 | 1.2 | 74 | 212 | 1.04 | 0.1 | 0.074 |
| v_zSTN|participant_id_sigma | 0.04 | 0.05 | 0.002 | 0.15 | 8 | 8 | 1.25 | 0.021 | 0.021 |
| v_zGPe|participant_id_sigma | 0.025 | 0.02 | 0.0023 | 0.057 | 103 | 193 | 1.04 | 0.0021 | 0.0035 |
| v_1|participant_id[1] | -0.002 | 0.037 | -0.053 | 0.057 | 298 | 595 | 1.01 | 0.0018 | 0.0022 |
| v_1|participant_id[2] | -0.004 | 0.035 | -0.061 | 0.047 | 383 | 269 | 1.01 | 0.0017 | 0.0019 |
| v_1|participant_id[3] | 0.028 | 0.044 | -0.018 | 0.11 | 363 | 594 | 1.02 | 0.0022 | 0.0025 |
| v_1|participant_id[4] | -0.011 | 0.033 | -0.065 | 0.033 | 729 | 584 | 1.01 | 0.0011 | 0.0013 |
| v_1|participant_id[5] | 0.005 | 0.032 | -0.043 | 0.062 | 966 | 961 | 1.02 | 0.00091 | 0.0011 |
| v_1|participant_id[6] | -0.02 | 0.04 | -0.1 | 0.026 | 53 | 80 | 1.04 | 0.0052 | 0.0053 |
| v_1|participant_id[7] | 0.003 | 0.035 | -0.052 | 0.06 | 899 | 705 | 1.02 | 0.0011 | 0.0015 |
| v_1|participant_id[8] | -0.015 | 0.04 | -0.083 | 0.035 | 87 | 127 | 1.03 | 0.0028 | 0.0036 |
| v_zGPe|participant_id_offset[1] | 0.3 | 0.88 | -1.1 | 1.6 | 219 | 522 | 1.02 | 0.061 | 0.045 |
| v_zGPe|participant_id_offset[2] | -0.23 | 0.81 | -1.6 | 0.99 | 1243 | 733 | 1.05 | 0.023 | 0.018 |
| v_zGPe|participant_id_offset[3] | 0.6 | 0.9 | -0.95 | 1.9 | 209 | 126 | 1.02 | 0.065 | 0.05 |
| v_zGPe|participant_id_offset[4] | 0.17 | 0.78 | -1.2 | 1.3 | 955 | 480 | 1.01 | 0.025 | 0.02 |
| v_zGPe|participant_id_offset[5] | -0.19 | 0.8 | -1.5 | 1.1 | 1181 | 865 | 1.04 | 0.023 | 0.019 |
| v_zGPe|participant_id_offset[6] | -0.38 | 0.85 | -1.8 | 1 | 617 | 478 | 1.02 | 0.035 | 0.028 |
| v_zGPe|participant_id_offset[7] | 0.38 | 0.82 | -1 | 1.6 | 453 | 492 | 1.01 | 0.04 | 0.031 |
| v_zGPe|participant_id_offset[8] | -0.4 | 0.8 | -1.8 | 0.84 | 119 | 303 | 1.03 | 0.076 | 0.058 |
| v_zGPe:sevScore | -0.187 | 0.051 | -0.25 | -0.1 | 158 | 261 | 1.02 | 0.0037 | 0.0043 |
| v_zGPe|participant_id[1] | 0.012 | 0.026 | -0.019 | 0.056 | 37 | 182 | 1.06 | 0.0037 | 0.0046 |
| v_zGPe|participant_id[2] | -0.004 | 0.019 | -0.036 | 0.022 | 686 | 550 | 1.03 | 0.00076 | 0.00088 |
| v_zGPe|participant_id[3] | 0.02 | 0.027 | -0.0096 | 0.069 | 98 | 86 | 1.04 | 0.0025 | 0.0029 |
| v_zGPe|participant_id[4] | 0.005 | 0.018 | -0.02 | 0.035 | 375 | 532 | 1.01 | 0.00086 | 0.00096 |
| v_zGPe|participant_id[5] | -0.005 | 0.018 | -0.036 | 0.019 | 449 | 835 | 1.01 | 0.00067 | 0.00071 |
| v_zGPe|participant_id[6] | -0.012 | 0.02 | -0.048 | 0.013 | 314 | 332 | 1.00 | 0.0011 | 0.0011 |
| v_zGPe|participant_id[7] | 0.01 | 0.02 | -0.017 | 0.044 | 33 | 65 | 1.07 | 0.0039 | 0.0033 |
| v_zGPe|participant_id[8] | -0.013 | 0.025 | -0.06 | 0.016 | 134 | 106 | 1.03 | 0.0019 | 0.0022 |
| v_zSTN | 0.81 | 0.06 | 0.7 | 0.87 | 8 | 5 | 1.28 | 0.025 | 0.018 |
| v_Intercept | 0.53 | 0.053 | 0.45 | 0.61 | 92 | 185 | 1.02 | 0.0054 | 0.004 |
| v_zSTN:sevScore | -0.4 | 0.2 | -0.54 | -0.061 | 8 | 6 | 1.27 | 0.073 | 0.06 |
| t | 0.5021 | 0.0069 | 0.49 | 0.51 | 324 | 503 | 1.01 | 0.00038 | 0.00029 |
| v_zSTN|participant_id_offset[1] | 0.2 | 0.9 | -1.4 | 1.6 | 32 | 259 | 1.07 | 0.16 | 0.12 |
| v_zSTN|participant_id_offset[2] | 0.14 | 0.84 | -1.3 | 1.4 | 470 | 731 | 1.05 | 0.042 | 0.03 |
| v_zSTN|participant_id_offset[3] | 0.17 | 0.81 | -1.2 | 1.4 | 1118 | 947 | 1.08 | 0.024 | 0.019 |
| v_zSTN|participant_id_offset[4] | -0.05 | 0.87 | -1.4 | 1.4 | 1794 | 843 | 1.10 | 0.02 | 0.017 |
| v_zSTN|participant_id_offset[5] | 0.3 | 0.8 | -1 | 1.6 | 138 | 721 | 1.02 | 0.069 | 0.052 |
| v_zSTN|participant_id_offset[6] | -0.3 | 0.8 | -1.5 | 1.1 | 138 | 508 | 1.03 | 0.073 | 0.057 |
| v_zSTN|participant_id_offset[7] | -0.1 | 0.9 | -1.4 | 1.4 | 25 | 129 | 1.08 | 0.19 | 0.12 |
| v_zSTN|participant_id_offset[8] | -0.8 | 1 | -2.2 | 0.89 | 17 | 57 | 1.13 | 0.24 | 0.16 |
| a | 1.474 | 0.02 | 1.4 | 1.5 | 23 | 72 | 1.09 | 0.004 | 0.0027 |
- there is no main effect of severity (makes sense because we didn't change the intercept by severity!)
- you can see that by looking at the credible interval (hdi's). They are interpret similar to confidence intervals
- the intercept itself is recovered well (input: 0.5)
- the effect of severity onto STN is significant. This is good because we did manipulate that (0.6 was around the mean effect)
Case Study 4: Two hierarchical layers (trial- and sbj-level) & Two within-sbj coefficients & one btw-sbj coefficient & group¶
- we still have the intracranial recordings
- but now you have different patients and they vary in symptom severity for Parkinson
- you wonder whether those with higher symptom severity have a lower drift rate because the modulation from STN is less efficient
BUT NOW: your friend is a clinician and tells you that some people might have Dystonia rather than Parkinson and they show different symptoms. So, you want to take this into account now
Audience Question: how would you do that? How would you change the data simulation?¶
Step 1: Regression-based data simulation (multiple subjects)¶
In [27]:
Copied!
# Function to simulate data for one participant
def simulate_participant3(participant_id, sevScore, diagnosis, size=300):
"""Simulate DDM trial data for one participant with diagnosis-dependent drift."""
# intercept = 0.5
zSTN = np.random.normal(loc=1, scale=2, size=size)
zGPe = np.random.normal(loc=1, scale=2, size=size)
if diagnosis == "PD":
intercept = 0.3
else:
intercept = 0.6
## Strength of neural modulation depends on PD severity score
if sevScore <= 0.2:
v = intercept + 0.8 * zSTN + 0.3 * zGPe
elif sevScore > 0.2 and sevScore <= 0.4:
v = intercept + 0.6 * zSTN + 0.2 * zGPe
elif sevScore > 0.4 and sevScore < 0.6:
v = intercept + 0.5 * zSTN + 0.1 * zGPe
else:
v = intercept + 0.4 * zSTN + 0.0 * zGPe
# Assume `hssm.simulate_data` returns a DataFrame
true_values = np.column_stack(
[v, np.repeat([[1.5, 0.5, 0.5]], axis=0, repeats=size)]
)
dataset_reg_v = hssm.simulate_data(
model="ddm",
theta=true_values,
size=1, # Generate one data point for each of the 1000 set of true values
)
# Adding additional variables to the dataset
dataset_reg_v["zSTN"] = zSTN
dataset_reg_v["zGPe"] = zGPe
dataset_reg_v["participant_id"] = str(participant_id)
dataset_reg_v["sevScore"] = sevScore
dataset_reg_v["diagnosis"] = diagnosis
return dataset_reg_v
# Simulate data for four participants
### note that we assume that STN & GPe
## PD patients:
# patients with low severity scores
subj_list = [1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8]
sevscore_list = [
0.10,
0.20,
0.30,
0.40,
0.50,
0.60,
0.70,
0.80,
0.10,
0.20,
0.30,
0.40,
0.50,
0.60,
0.70,
0.80,
]
diagnosis_list = [
"PD",
"PD",
"PD",
"PD",
"PD",
"PD",
"PD",
"PD",
"DD",
"DD",
"DD",
"DD",
"DD",
"DD",
"DD",
"DD",
]
# [simulate_participant3(subj, sevscore, diagnosis) for subj, sevscore, diagnosis
# in zip(subj_list, sevscore_list, diagnosis_list)]
# dataset2_participant1 = simulate_participant3(1,0.10,"PD")
# dataset2_participant2 = simulate_participant3(2,0.20,"PD")
# dataset2_participant3 = simulate_participant3(3,0.30,"PD")
# dataset2_participant4 = simulate_participant3(4,0.40,"PD")
# # patients with high severity scores
# dataset2_participant5 = simulate_participant3(5,0.50,"PD")
# dataset2_participant6 = simulate_participant3(6,0.60,"PD")
# dataset2_participant7 = simulate_participant3(7,0.70,"PD")
# dataset2_participant8 = simulate_participant3(8,0.80,"PD")
# ## Dystonia patients:
# # patients with low severity scores
# dataset2_participant9 = simulate_participant3(1,0.10,"DD")
# dataset2_participant10 = simulate_participant3(2,0.20,"DD")
# dataset2_participant11 = simulate_participant3(3,0.30,"DD")
# dataset2_participant12 = simulate_participant3(4,0.40,"DD")
# # patients with high severity scores
# dataset2_participant13 = simulate_participant3(5,0.50,"DD")
# dataset2_participant14 = simulate_participant3(6,0.60,"DD")
# dataset2_participant15 = simulate_participant3(7,0.70,"DD")
# dataset2_participant16 = simulate_participant3(8,0.80,"DD")
# Combine datasets into one DataFrame
combined_dataset3 = pd.concat(
[
simulate_participant3(subj, sevscore, diagnosis)
for subj, sevscore, diagnosis in zip(subj_list, sevscore_list, diagnosis_list)
],
ignore_index=True,
)
combined_dataset3
# Function to simulate data for one participant
def simulate_participant3(participant_id, sevScore, diagnosis, size=300):
"""Simulate DDM trial data for one participant with diagnosis-dependent drift."""
# intercept = 0.5
zSTN = np.random.normal(loc=1, scale=2, size=size)
zGPe = np.random.normal(loc=1, scale=2, size=size)
if diagnosis == "PD":
intercept = 0.3
else:
intercept = 0.6
## Strength of neural modulation depends on PD severity score
if sevScore <= 0.2:
v = intercept + 0.8 * zSTN + 0.3 * zGPe
elif sevScore > 0.2 and sevScore <= 0.4:
v = intercept + 0.6 * zSTN + 0.2 * zGPe
elif sevScore > 0.4 and sevScore < 0.6:
v = intercept + 0.5 * zSTN + 0.1 * zGPe
else:
v = intercept + 0.4 * zSTN + 0.0 * zGPe
# Assume `hssm.simulate_data` returns a DataFrame
true_values = np.column_stack(
[v, np.repeat([[1.5, 0.5, 0.5]], axis=0, repeats=size)]
)
dataset_reg_v = hssm.simulate_data(
model="ddm",
theta=true_values,
size=1, # Generate one data point for each of the 1000 set of true values
)
# Adding additional variables to the dataset
dataset_reg_v["zSTN"] = zSTN
dataset_reg_v["zGPe"] = zGPe
dataset_reg_v["participant_id"] = str(participant_id)
dataset_reg_v["sevScore"] = sevScore
dataset_reg_v["diagnosis"] = diagnosis
return dataset_reg_v
# Simulate data for four participants
### note that we assume that STN & GPe
## PD patients:
# patients with low severity scores
subj_list = [1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 3, 4, 5, 6, 7, 8]
sevscore_list = [
0.10,
0.20,
0.30,
0.40,
0.50,
0.60,
0.70,
0.80,
0.10,
0.20,
0.30,
0.40,
0.50,
0.60,
0.70,
0.80,
]
diagnosis_list = [
"PD",
"PD",
"PD",
"PD",
"PD",
"PD",
"PD",
"PD",
"DD",
"DD",
"DD",
"DD",
"DD",
"DD",
"DD",
"DD",
]
# [simulate_participant3(subj, sevscore, diagnosis) for subj, sevscore, diagnosis
# in zip(subj_list, sevscore_list, diagnosis_list)]
# dataset2_participant1 = simulate_participant3(1,0.10,"PD")
# dataset2_participant2 = simulate_participant3(2,0.20,"PD")
# dataset2_participant3 = simulate_participant3(3,0.30,"PD")
# dataset2_participant4 = simulate_participant3(4,0.40,"PD")
# # patients with high severity scores
# dataset2_participant5 = simulate_participant3(5,0.50,"PD")
# dataset2_participant6 = simulate_participant3(6,0.60,"PD")
# dataset2_participant7 = simulate_participant3(7,0.70,"PD")
# dataset2_participant8 = simulate_participant3(8,0.80,"PD")
# ## Dystonia patients:
# # patients with low severity scores
# dataset2_participant9 = simulate_participant3(1,0.10,"DD")
# dataset2_participant10 = simulate_participant3(2,0.20,"DD")
# dataset2_participant11 = simulate_participant3(3,0.30,"DD")
# dataset2_participant12 = simulate_participant3(4,0.40,"DD")
# # patients with high severity scores
# dataset2_participant13 = simulate_participant3(5,0.50,"DD")
# dataset2_participant14 = simulate_participant3(6,0.60,"DD")
# dataset2_participant15 = simulate_participant3(7,0.70,"DD")
# dataset2_participant16 = simulate_participant3(8,0.80,"DD")
# Combine datasets into one DataFrame
combined_dataset3 = pd.concat(
[
simulate_participant3(subj, sevscore, diagnosis)
for subj, sevscore, diagnosis in zip(subj_list, sevscore_list, diagnosis_list)
],
ignore_index=True,
)
combined_dataset3
Out[27]:
| rt | response | zSTN | zGPe | participant_id | sevScore | diagnosis | |
|---|---|---|---|---|---|---|---|
| 0 | 0.857299 | 1.0 | 4.174684 | 1.511771 | 1 | 0.1 | PD |
| 1 | 0.986127 | 1.0 | 3.633133 | -0.442452 | 1 | 0.1 | PD |
| 2 | 1.410487 | 1.0 | 1.415882 | 1.524683 | 1 | 0.1 | PD |
| 3 | 1.979731 | 1.0 | 0.930716 | -0.456628 | 1 | 0.1 | PD |
| 4 | 0.721911 | 1.0 | 4.502797 | 2.518944 | 1 | 0.1 | PD |
| ... | ... | ... | ... | ... | ... | ... | ... |
| 4795 | 1.022368 | 1.0 | 4.160187 | 5.785448 | 8 | 0.8 | DD |
| 4796 | 1.241501 | -1.0 | 0.315886 | -0.101750 | 8 | 0.8 | DD |
| 4797 | 1.106278 | 1.0 | 5.438028 | -0.364264 | 8 | 0.8 | DD |
| 4798 | 1.351487 | 1.0 | 0.714511 | 2.704185 | 8 | 0.8 | DD |
| 4799 | 1.178348 | 1.0 | 5.107501 | 2.936340 | 8 | 0.8 | DD |
4800 rows × 7 columns
In [28]:
Copied!
model_reg_v_ex4_A1 = hssm.HSSM(
data=combined_dataset3,
include=[
{
"name": "v",
"formula": (
"v ~ 1 + (zSTN + zGPe)*sevScore*C(diagnosis) + "
"((1 + zSTN + zGPe)|participant_id)"
),
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.5},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zSTN:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
model_reg_v_ex4_A1
model_reg_v_ex4_A1 = hssm.HSSM(
data=combined_dataset3,
include=[
{
"name": "v",
"formula": (
"v ~ 1 + (zSTN + zGPe)*sevScore*C(diagnosis) + "
"((1 + zSTN + zGPe)|participant_id)"
),
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.5},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zSTN:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
model_reg_v_ex4_A1
Model initialized successfully.
Out[28]:
Hierarchical Sequential Sampling Model
Model: ddm
Response variable: rt,response
Likelihood: analytical
Observations: 4800
Parameters:
v:
Formula: v ~ 1 + (zSTN + zGPe)*sevScore*C(diagnosis) + ((1 + zSTN + zGPe)|participant_id)
Priors:
v_Intercept ~ Normal(mu: 1.5, sigma: 1.5)
v_zSTN ~ Normal(mu: 0.0, sigma: 1.0)
v_zGPe ~ Normal(mu: 0.0, sigma: 1.0)
v_sevScore ~ Normal(mu: 0.0, sigma: 1.0)
v_zSTN:sevScore ~ Normal(mu: 0.0, sigma: 1.0)
v_zGPe:sevScore ~ Normal(mu: 0.0, sigma: 1.0)
v_C(diagnosis) ~ Normal(mu: 0.0, sigma: 0.25)
v_zSTN:C(diagnosis) ~ Normal(mu: 0.0, sigma: 0.25)
v_zGPe:C(diagnosis) ~ Normal(mu: 0.0, sigma: 0.25)
v_sevScore:C(diagnosis) ~ Normal(mu: 0.0, sigma: 0.25)
v_zSTN:sevScore:C(diagnosis) ~ Normal(mu: 0.0, sigma: 0.25)
v_zGPe:sevScore:C(diagnosis) ~ Normal(mu: 0.0, sigma: 0.25)
v_1|participant_id ~ Normal(mu: 0.0, sigma: HalfNormal(sigma: 1.0))
v_zSTN|participant_id ~ Normal(mu: 0.0, sigma: HalfNormal(sigma: 1.0))
v_zGPe|participant_id ~ Normal(mu: 0.0, sigma: HalfNormal(sigma: 1.0))
Link: identity
Explicit bounds: (-inf, inf)
a:
Prior: HalfNormal(sigma: 2.0)
Explicit bounds: (0.0, inf)
z:
Prior: Uniform(lower: 0.0, upper: 1.0)
Explicit bounds: (0.0, 1.0)
t:
Prior: HalfNormal(sigma: 2.0)
Explicit bounds: (0.0, inf)
Lapse probability: 0.05
Lapse distribution: Uniform(lower: 0.0, upper: 20.0)
In [29]:
Copied!
samples_model_reg_v_ex4_A1 = model_reg_v_ex4_A1.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
samples_model_reg_v_ex4_A1 = model_reg_v_ex4_A1.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
Using default initvals.
NUTS[numpyro]: [a, z, t, v_Intercept, v_zSTN, v_zGPe, v_sevScore, v_zSTN:sevScore, v_zGPe:sevScore, v_C(diagnosis), v_zSTN:C(diagnosis), v_zGPe:C(diagnosis), v_sevScore:C(diagnosis), v_zSTN:sevScore:C(diagnosis), v_zGPe:sevScore:C(diagnosis), v_1|participant_id_sigma, v_1|participant_id_offset, v_zSTN|participant_id_sigma, v_zSTN|participant_id_offset, v_zGPe|participant_id_sigma, v_zGPe|participant_id_offset]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 0%| | 0/1000 [00:10<?, ?it/s, 1023 steps of size 2.14e-03. acc. prob=0.69]
warmup: 3%|▎ | 28/1000 [00:10<06:06, 2.65it/s, 1023 steps of size 3.82e-03. acc. prob=0.70]
warmup: 3%|▎ | 29/1000 [00:11<06:24, 2.53it/s, 1023 steps of size 5.40e-03. acc. prob=0.71]
warmup: 3%|▎ | 30/1000 [00:11<06:21, 2.54it/s, 511 steps of size 1.92e-03. acc. prob=0.70]
warmup: 3%|▎ | 31/1000 [00:12<06:46, 2.38it/s, 1023 steps of size 3.49e-03. acc. prob=0.71]
warmup: 3%|▎ | 32/1000 [00:12<06:40, 2.42it/s, 511 steps of size 6.31e-03. acc. prob=0.71]
warmup: 3%|▎ | 33/1000 [00:13<06:33, 2.46it/s, 511 steps of size 6.04e-03. acc. prob=0.72]
warmup: 3%|▎ | 34/1000 [00:13<06:25, 2.51it/s, 511 steps of size 9.74e-03. acc. prob=0.72]
warmup: 4%|▎ | 36/1000 [00:14<06:12, 2.59it/s, 1023 steps of size 2.60e-03. acc. prob=0.71]
warmup: 4%|▎ | 37/1000 [00:14<07:09, 2.24it/s, 1023 steps of size 4.13e-03. acc. prob=0.72]
warmup: 4%|▍ | 38/1000 [00:15<06:51, 2.34it/s, 511 steps of size 6.74e-03. acc. prob=0.73]
warmup: 4%|▍ | 39/1000 [00:15<06:36, 2.42it/s, 511 steps of size 8.08e-03. acc. prob=0.73]
warmup: 4%|▍ | 40/1000 [00:15<06:26, 2.48it/s, 511 steps of size 8.53e-03. acc. prob=0.73]
warmup: 4%|▍ | 41/1000 [00:16<05:31, 2.89it/s, 255 steps of size 3.24e-03. acc. prob=0.72]
warmup: 4%|▍ | 42/1000 [00:16<07:06, 2.24it/s, 1023 steps of size 4.19e-03. acc. prob=0.73]
warmup: 4%|▍ | 43/1000 [00:17<08:24, 1.90it/s, 1023 steps of size 7.21e-03. acc. prob=0.73]
warmup: 4%|▍ | 44/1000 [00:18<07:58, 2.00it/s, 511 steps of size 7.11e-03. acc. prob=0.73]
warmup: 4%|▍ | 45/1000 [00:18<06:28, 2.46it/s, 255 steps of size 4.31e-03. acc. prob=0.73]
warmup: 5%|▍ | 46/1000 [00:18<07:55, 2.01it/s, 1023 steps of size 6.05e-03. acc. prob=0.73]
warmup: 5%|▍ | 47/1000 [00:19<07:18, 2.17it/s, 511 steps of size 5.62e-03. acc. prob=0.73]
warmup: 5%|▍ | 48/1000 [00:19<06:50, 2.32it/s, 511 steps of size 9.43e-03. acc. prob=0.74]
warmup: 5%|▌ | 50/1000 [00:20<06:19, 2.50it/s, 1023 steps of size 2.97e-03. acc. prob=0.73]
warmup: 5%|▌ | 51/1000 [00:21<07:33, 2.09it/s, 1023 steps of size 4.58e-03. acc. prob=0.74]
warmup: 5%|▌ | 52/1000 [00:21<07:05, 2.23it/s, 511 steps of size 7.33e-03. acc. prob=0.74]
warmup: 5%|▌ | 53/1000 [00:21<06:41, 2.36it/s, 511 steps of size 3.01e-03. acc. prob=0.73]
warmup: 5%|▌ | 54/1000 [00:22<08:00, 1.97it/s, 1023 steps of size 4.88e-03. acc. prob=0.74]
warmup: 6%|▌ | 55/1000 [00:23<08:56, 1.76it/s, 1023 steps of size 4.94e-03. acc. prob=0.74]
warmup: 6%|▌ | 56/1000 [00:23<07:57, 1.98it/s, 511 steps of size 7.74e-03. acc. prob=0.74]
warmup: 6%|▌ | 57/1000 [00:23<06:27, 2.43it/s, 255 steps of size 6.92e-03. acc. prob=0.74]
warmup: 6%|▌ | 58/1000 [00:24<06:11, 2.53it/s, 511 steps of size 5.11e-03. acc. prob=0.74]
warmup: 6%|▌ | 59/1000 [00:24<06:00, 2.61it/s, 511 steps of size 8.34e-03. acc. prob=0.75]
warmup: 6%|▌ | 60/1000 [00:24<05:54, 2.65it/s, 511 steps of size 1.71e-03. acc. prob=0.73]
warmup: 6%|▌ | 61/1000 [00:25<07:38, 2.05it/s, 1023 steps of size 2.79e-03. acc. prob=0.74]
warmup: 6%|▌ | 62/1000 [00:26<08:40, 1.80it/s, 1023 steps of size 4.45e-03. acc. prob=0.74]
warmup: 6%|▋ | 63/1000 [00:27<09:27, 1.65it/s, 1023 steps of size 7.19e-03. acc. prob=0.75]
warmup: 6%|▋ | 64/1000 [00:27<08:17, 1.88it/s, 511 steps of size 4.54e-03. acc. prob=0.74]
warmup: 6%|▋ | 65/1000 [00:27<07:30, 2.08it/s, 511 steps of size 6.00e-03. acc. prob=0.75]
warmup: 7%|▋ | 66/1000 [00:28<06:55, 2.25it/s, 511 steps of size 6.41e-03. acc. prob=0.75]
warmup: 7%|▋ | 67/1000 [00:28<06:31, 2.38it/s, 511 steps of size 4.21e-03. acc. prob=0.75]
warmup: 7%|▋ | 68/1000 [00:29<07:52, 1.97it/s, 1023 steps of size 5.19e-03. acc. prob=0.75]
warmup: 7%|▋ | 69/1000 [00:29<08:51, 1.75it/s, 1023 steps of size 4.56e-03. acc. prob=0.75]
warmup: 7%|▋ | 70/1000 [00:30<07:51, 1.97it/s, 511 steps of size 5.75e-03. acc. prob=0.75]
warmup: 7%|▋ | 71/1000 [00:30<07:09, 2.16it/s, 511 steps of size 4.47e-03. acc. prob=0.75]
warmup: 7%|▋ | 72/1000 [00:31<06:39, 2.32it/s, 511 steps of size 3.17e-03. acc. prob=0.75]
warmup: 7%|▋ | 73/1000 [00:31<07:58, 1.94it/s, 1023 steps of size 4.60e-03. acc. prob=0.75]
warmup: 7%|▋ | 74/1000 [00:32<08:52, 1.74it/s, 1023 steps of size 3.93e-03. acc. prob=0.75]
warmup: 8%|▊ | 75/1000 [00:33<09:31, 1.62it/s, 1023 steps of size 5.10e-03. acc. prob=0.75]
warmup: 8%|▊ | 76/1000 [00:33<08:19, 1.85it/s, 511 steps of size 5.29e-03. acc. prob=0.75]
warmup: 8%|▊ | 77/1000 [00:34<08:15, 1.86it/s, 767 steps of size 8.20e-03. acc. prob=0.75]
warmup: 8%|▊ | 78/1000 [00:34<06:36, 2.33it/s, 255 steps of size 6.04e-03. acc. prob=0.75]
warmup: 8%|▊ | 79/1000 [00:34<06:14, 2.46it/s, 511 steps of size 7.00e-03. acc. prob=0.75]
warmup: 8%|▊ | 80/1000 [00:34<06:02, 2.54it/s, 511 steps of size 7.71e-03. acc. prob=0.75]
warmup: 8%|▊ | 81/1000 [00:35<04:57, 3.09it/s, 223 steps of size 3.81e-03. acc. prob=0.75]
warmup: 8%|▊ | 82/1000 [00:35<06:45, 2.26it/s, 1023 steps of size 5.62e-03. acc. prob=0.75]
warmup: 8%|▊ | 83/1000 [00:36<06:22, 2.39it/s, 511 steps of size 7.12e-03. acc. prob=0.76]
warmup: 8%|▊ | 84/1000 [00:36<06:05, 2.50it/s, 511 steps of size 6.92e-03. acc. prob=0.76]
warmup: 8%|▊ | 85/1000 [00:36<05:54, 2.58it/s, 511 steps of size 5.45e-03. acc. prob=0.75]
warmup: 9%|▊ | 86/1000 [00:37<07:24, 2.06it/s, 1023 steps of size 7.90e-03. acc. prob=0.76]
warmup: 9%|▊ | 87/1000 [00:38<06:49, 2.23it/s, 511 steps of size 9.24e-03. acc. prob=0.76]
warmup: 9%|▉ | 89/1000 [00:38<06:11, 2.45it/s, 1023 steps of size 3.46e-03. acc. prob=0.75]
warmup: 9%|▉ | 90/1000 [00:39<07:20, 2.06it/s, 1023 steps of size 4.58e-03. acc. prob=0.75]
warmup: 9%|▉ | 91/1000 [00:40<08:14, 1.84it/s, 1023 steps of size 5.23e-03. acc. prob=0.76]
warmup: 9%|▉ | 92/1000 [00:40<07:26, 2.03it/s, 511 steps of size 7.84e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:40<06:51, 2.21it/s, 511 steps of size 3.50e-03. acc. prob=0.75]
warmup: 9%|▉ | 94/1000 [00:41<07:58, 1.89it/s, 1023 steps of size 5.11e-03. acc. prob=0.76]
warmup: 10%|▉ | 95/1000 [00:42<08:47, 1.72it/s, 1023 steps of size 6.37e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:42<07:46, 1.94it/s, 511 steps of size 9.33e-03. acc. prob=0.76]
warmup: 10%|▉ | 98/1000 [00:43<06:42, 2.24it/s, 1023 steps of size 4.82e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:43<06:22, 2.35it/s, 511 steps of size 4.79e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:44<07:28, 2.01it/s, 1023 steps of size 5.70e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:45<08:21, 1.79it/s, 1023 steps of size 8.20e-02. acc. prob=0.76]
warmup: 10%|█ | 104/1000 [00:45<04:09, 3.59it/s, 31 steps of size 9.25e-02. acc. prob=0.76]
warmup: 11%|█ | 107/1000 [00:45<02:35, 5.74it/s, 31 steps of size 1.89e-01. acc. prob=0.77]
warmup: 11%|█ | 109/1000 [00:45<02:13, 6.66it/s, 255 steps of size 2.92e-02. acc. prob=0.76]
warmup: 11%|█ | 111/1000 [00:45<01:57, 7.54it/s, 127 steps of size 7.54e-02. acc. prob=0.76]
warmup: 12%|█▏ | 115/1000 [00:46<01:31, 9.72it/s, 255 steps of size 3.11e-02. acc. prob=0.76]
warmup: 12%|█▏ | 117/1000 [00:46<01:28, 10.03it/s, 127 steps of size 9.25e-02. acc. prob=0.77]
warmup: 12%|█▏ | 119/1000 [00:46<01:20, 10.95it/s, 63 steps of size 1.78e-01. acc. prob=0.77]
warmup: 12%|█▏ | 121/1000 [00:46<01:20, 10.91it/s, 255 steps of size 4.90e-02. acc. prob=0.77]
warmup: 12%|█▏ | 123/1000 [00:46<01:11, 12.29it/s, 31 steps of size 1.13e-01. acc. prob=0.77]
warmup: 13%|█▎ | 126/1000 [00:46<00:59, 14.67it/s, 127 steps of size 5.82e-02. acc. prob=0.77]
warmup: 13%|█▎ | 128/1000 [00:46<00:59, 14.69it/s, 63 steps of size 1.52e-01. acc. prob=0.77]
warmup: 13%|█▎ | 130/1000 [00:47<00:56, 15.41it/s, 127 steps of size 5.54e-02. acc. prob=0.77]
warmup: 13%|█▎ | 132/1000 [00:47<00:57, 15.21it/s, 63 steps of size 8.77e-02. acc. prob=0.77]
warmup: 13%|█▎ | 134/1000 [00:47<01:02, 13.89it/s, 127 steps of size 1.13e-01. acc. prob=0.77]
warmup: 14%|█▎ | 137/1000 [00:47<01:03, 13.53it/s, 255 steps of size 4.09e-02. acc. prob=0.77]
warmup: 14%|█▍ | 139/1000 [00:47<01:07, 12.76it/s, 127 steps of size 9.31e-02. acc. prob=0.77]
warmup: 14%|█▍ | 142/1000 [00:47<01:07, 12.74it/s, 255 steps of size 3.38e-02. acc. prob=0.77]
warmup: 14%|█▍ | 143/1000 [00:48<01:20, 10.66it/s, 255 steps of size 5.43e-02. acc. prob=0.77]
warmup: 14%|█▍ | 145/1000 [00:48<01:13, 11.63it/s, 63 steps of size 1.14e-01. acc. prob=0.77]
warmup: 15%|█▍ | 149/1000 [00:48<00:57, 14.92it/s, 127 steps of size 6.49e-02. acc. prob=0.77]
warmup: 15%|█▌ | 153/1000 [00:48<00:45, 18.56it/s, 63 steps of size 6.96e-02. acc. prob=0.77]
warmup: 16%|█▌ | 157/1000 [00:48<00:49, 17.12it/s, 255 steps of size 2.15e-02. acc. prob=0.77]
warmup: 16%|█▌ | 159/1000 [00:49<00:54, 15.45it/s, 127 steps of size 6.42e-02. acc. prob=0.77]
warmup: 16%|█▌ | 161/1000 [00:49<00:59, 14.19it/s, 127 steps of size 4.93e-02. acc. prob=0.77]
warmup: 16%|█▋ | 164/1000 [00:49<00:51, 16.09it/s, 63 steps of size 1.27e-01. acc. prob=0.77]
warmup: 17%|█▋ | 166/1000 [00:49<01:19, 10.46it/s, 511 steps of size 3.11e-02. acc. prob=0.77]
warmup: 17%|█▋ | 168/1000 [00:49<01:13, 11.31it/s, 63 steps of size 1.04e-01. acc. prob=0.77]
warmup: 17%|█▋ | 172/1000 [00:50<01:05, 12.55it/s, 255 steps of size 3.72e-02. acc. prob=0.77]
warmup: 17%|█▋ | 173/1000 [00:50<01:16, 10.79it/s, 255 steps of size 6.67e-02. acc. prob=0.77]
warmup: 18%|█▊ | 175/1000 [00:50<01:08, 12.08it/s, 31 steps of size 1.25e-01. acc. prob=0.77]
warmup: 18%|█▊ | 177/1000 [00:50<01:32, 8.90it/s, 511 steps of size 4.06e-02. acc. prob=0.77]
warmup: 18%|█▊ | 179/1000 [00:50<01:21, 10.07it/s, 63 steps of size 1.10e-01. acc. prob=0.77]
warmup: 18%|█▊ | 182/1000 [00:51<01:04, 12.74it/s, 127 steps of size 4.97e-02. acc. prob=0.77]
warmup: 18%|█▊ | 184/1000 [00:51<01:01, 13.17it/s, 127 steps of size 7.44e-02. acc. prob=0.77]
warmup: 19%|█▉ | 188/1000 [00:51<00:57, 14.22it/s, 255 steps of size 4.64e-02. acc. prob=0.77]
warmup: 19%|█▉ | 190/1000 [00:51<01:00, 13.37it/s, 127 steps of size 4.68e-02. acc. prob=0.77]
warmup: 19%|█▉ | 192/1000 [00:51<00:56, 14.29it/s, 31 steps of size 1.22e-01. acc. prob=0.78]
warmup: 19%|█▉ | 194/1000 [00:51<00:54, 14.73it/s, 127 steps of size 8.61e-02. acc. prob=0.77]
warmup: 20%|█▉ | 198/1000 [00:52<00:47, 16.93it/s, 127 steps of size 8.51e-02. acc. prob=0.78]
warmup: 20%|██ | 200/1000 [00:52<00:48, 16.42it/s, 127 steps of size 7.82e-02. acc. prob=0.78]
warmup: 20%|██ | 203/1000 [00:52<00:44, 18.09it/s, 63 steps of size 1.36e-01. acc. prob=0.78]
warmup: 20%|██ | 205/1000 [00:52<00:50, 15.71it/s, 255 steps of size 4.05e-02. acc. prob=0.77]
warmup: 21%|██ | 207/1000 [00:52<00:55, 14.28it/s, 127 steps of size 7.95e-02. acc. prob=0.78]
warmup: 21%|██ | 210/1000 [00:52<00:51, 15.23it/s, 127 steps of size 5.95e-02. acc. prob=0.78]
warmup: 21%|██▏ | 213/1000 [00:53<00:48, 16.34it/s, 127 steps of size 5.69e-02. acc. prob=0.78]
warmup: 22%|██▏ | 216/1000 [00:53<00:41, 19.09it/s, 15 steps of size 2.81e-02. acc. prob=0.77]
warmup: 22%|██▏ | 218/1000 [00:53<00:48, 16.28it/s, 127 steps of size 6.66e-02. acc. prob=0.78]
warmup: 22%|██▏ | 222/1000 [00:53<00:39, 19.77it/s, 63 steps of size 5.21e-02. acc. prob=0.78]
warmup: 22%|██▏ | 224/1000 [00:53<00:42, 18.41it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 23%|██▎ | 227/1000 [00:53<00:40, 19.31it/s, 127 steps of size 5.48e-02. acc. prob=0.78]
warmup: 23%|██▎ | 228/1000 [00:53<00:53, 14.47it/s, 255 steps of size 8.30e-02. acc. prob=0.78]
warmup: 23%|██▎ | 231/1000 [00:54<00:46, 16.64it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 24%|██▎ | 235/1000 [00:54<00:46, 16.51it/s, 255 steps of size 3.94e-02. acc. prob=0.78]
warmup: 24%|██▎ | 237/1000 [00:54<00:47, 16.20it/s, 63 steps of size 8.28e-02. acc. prob=0.78]
warmup: 24%|██▍ | 240/1000 [00:54<00:40, 18.67it/s, 63 steps of size 1.08e-01. acc. prob=0.78]
warmup: 24%|██▍ | 243/1000 [00:54<00:36, 20.65it/s, 31 steps of size 4.61e-02. acc. prob=0.78]
warmup: 24%|██▍ | 245/1000 [00:54<00:39, 18.91it/s, 63 steps of size 8.55e-02. acc. prob=0.78]
warmup: 25%|██▍ | 248/1000 [00:54<00:41, 18.24it/s, 127 steps of size 7.33e-02. acc. prob=0.78]
warmup: 25%|██▌ | 251/1000 [00:55<00:42, 17.66it/s, 127 steps of size 9.31e-01. acc. prob=0.78]
warmup: 25%|██▌ | 254/1000 [00:55<00:43, 17.12it/s, 255 steps of size 2.47e-02. acc. prob=0.77]
warmup: 26%|██▌ | 256/1000 [00:55<00:48, 15.19it/s, 127 steps of size 5.91e-02. acc. prob=0.78]
warmup: 26%|██▌ | 259/1000 [00:55<00:55, 13.41it/s, 255 steps of size 4.99e-02. acc. prob=0.78]
warmup: 26%|██▌ | 261/1000 [00:55<00:53, 13.70it/s, 63 steps of size 3.78e-02. acc. prob=0.78]
warmup: 26%|██▋ | 263/1000 [00:56<00:52, 14.04it/s, 63 steps of size 9.63e-02. acc. prob=0.78]
warmup: 27%|██▋ | 266/1000 [00:56<00:45, 16.10it/s, 63 steps of size 1.66e-01. acc. prob=0.78]
warmup: 27%|██▋ | 268/1000 [00:56<00:52, 13.86it/s, 255 steps of size 5.07e-02. acc. prob=0.78]
warmup: 27%|██▋ | 269/1000 [00:56<00:56, 13.01it/s, 127 steps of size 9.29e-02. acc. prob=0.78]
warmup: 27%|██▋ | 270/1000 [00:56<00:59, 12.22it/s, 127 steps of size 1.33e-01. acc. prob=0.78]
warmup: 27%|██▋ | 272/1000 [00:56<00:54, 13.38it/s, 127 steps of size 7.65e-02. acc. prob=0.78]
warmup: 28%|██▊ | 276/1000 [00:56<00:43, 16.65it/s, 127 steps of size 4.20e-02. acc. prob=0.78]
warmup: 28%|██▊ | 277/1000 [00:57<00:52, 13.64it/s, 191 steps of size 6.97e-02. acc. prob=0.78]
warmup: 28%|██▊ | 280/1000 [00:57<00:43, 16.70it/s, 24 steps of size 1.84e-02. acc. prob=0.78]
warmup: 28%|██▊ | 281/1000 [00:57<00:58, 12.28it/s, 255 steps of size 3.25e-02. acc. prob=0.78]
warmup: 28%|██▊ | 283/1000 [00:57<01:02, 11.55it/s, 127 steps of size 9.35e-02. acc. prob=0.78]
warmup: 29%|██▊ | 286/1000 [00:57<00:53, 13.45it/s, 127 steps of size 7.45e-02. acc. prob=0.78]
warmup: 29%|██▉ | 290/1000 [00:58<00:51, 13.77it/s, 255 steps of size 5.92e-02. acc. prob=0.78]
warmup: 29%|██▉ | 293/1000 [00:58<00:45, 15.61it/s, 63 steps of size 1.53e-01. acc. prob=0.78]
warmup: 30%|██▉ | 295/1000 [00:58<00:43, 16.07it/s, 127 steps of size 6.20e-02. acc. prob=0.78]
warmup: 30%|██▉ | 298/1000 [00:58<00:39, 17.70it/s, 63 steps of size 1.40e-01. acc. prob=0.78]
warmup: 30%|███ | 301/1000 [00:58<00:43, 16.02it/s, 255 steps of size 4.07e-02. acc. prob=0.78]
warmup: 30%|███ | 303/1000 [00:58<00:51, 13.57it/s, 191 steps of size 8.93e-02. acc. prob=0.78]
warmup: 30%|███ | 305/1000 [00:58<00:50, 13.89it/s, 127 steps of size 1.50e-01. acc. prob=0.78]
warmup: 31%|███ | 307/1000 [00:59<00:53, 12.93it/s, 255 steps of size 4.27e-02. acc. prob=0.78]
warmup: 31%|███ | 309/1000 [00:59<00:51, 13.41it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 31%|███▏ | 313/1000 [00:59<00:42, 16.32it/s, 127 steps of size 7.83e-02. acc. prob=0.78]
warmup: 32%|███▏ | 317/1000 [00:59<00:37, 18.28it/s, 127 steps of size 8.90e-02. acc. prob=0.78]
warmup: 32%|███▏ | 321/1000 [00:59<00:30, 22.14it/s, 31 steps of size 5.31e-02. acc. prob=0.78]
warmup: 32%|███▏ | 323/1000 [00:59<00:33, 20.36it/s, 127 steps of size 9.80e-02. acc. prob=0.78]
warmup: 33%|███▎ | 326/1000 [01:00<00:30, 22.13it/s, 63 steps of size 1.15e-01. acc. prob=0.78]
warmup: 33%|███▎ | 329/1000 [01:00<00:28, 23.51it/s, 63 steps of size 9.83e-02. acc. prob=0.78]
warmup: 33%|███▎ | 332/1000 [01:00<00:30, 22.01it/s, 127 steps of size 6.71e-02. acc. prob=0.78]
warmup: 34%|███▎ | 335/1000 [01:00<00:28, 23.37it/s, 31 steps of size 4.61e-02. acc. prob=0.78]
warmup: 34%|███▎ | 337/1000 [01:00<00:31, 20.86it/s, 63 steps of size 9.66e-02. acc. prob=0.78]
warmup: 34%|███▍ | 340/1000 [01:00<00:30, 21.55it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 34%|███▍ | 343/1000 [01:00<00:30, 21.80it/s, 63 steps of size 1.00e-01. acc. prob=0.78]
warmup: 35%|███▍ | 346/1000 [01:01<00:37, 17.64it/s, 255 steps of size 6.03e-02. acc. prob=0.78]
warmup: 35%|███▍ | 349/1000 [01:01<00:33, 19.72it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 35%|███▌ | 353/1000 [01:01<00:36, 17.59it/s, 255 steps of size 5.01e-02. acc. prob=0.78]
warmup: 36%|███▌ | 356/1000 [01:01<00:34, 18.76it/s, 63 steps of size 1.19e-01. acc. prob=0.78]
warmup: 36%|███▌ | 359/1000 [01:01<00:31, 20.34it/s, 127 steps of size 5.87e-02. acc. prob=0.78]
warmup: 36%|███▌ | 361/1000 [01:01<00:36, 17.29it/s, 127 steps of size 9.43e-02. acc. prob=0.78]
warmup: 36%|███▋ | 364/1000 [01:01<00:32, 19.37it/s, 31 steps of size 9.12e-02. acc. prob=0.78]
warmup: 37%|███▋ | 367/1000 [01:02<00:32, 19.61it/s, 127 steps of size 9.16e-02. acc. prob=0.78]
warmup: 37%|███▋ | 370/1000 [01:02<00:30, 20.48it/s, 63 steps of size 1.49e-01. acc. prob=0.78]
warmup: 37%|███▋ | 372/1000 [01:02<00:31, 19.83it/s, 127 steps of size 5.37e-02. acc. prob=0.78]
warmup: 37%|███▋ | 374/1000 [01:02<00:37, 16.72it/s, 127 steps of size 9.41e-02. acc. prob=0.78]
warmup: 38%|███▊ | 377/1000 [01:02<00:32, 19.25it/s, 31 steps of size 1.11e-01. acc. prob=0.78]
warmup: 38%|███▊ | 380/1000 [01:02<00:32, 19.29it/s, 127 steps of size 7.92e-02. acc. prob=0.78]
warmup: 38%|███▊ | 383/1000 [01:02<00:28, 21.31it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 39%|███▊ | 386/1000 [01:03<00:26, 22.90it/s, 63 steps of size 1.29e-01. acc. prob=0.78]
warmup: 39%|███▉ | 390/1000 [01:03<00:23, 26.46it/s, 63 steps of size 8.47e-02. acc. prob=0.78]
warmup: 39%|███▉ | 393/1000 [01:03<00:22, 26.51it/s, 63 steps of size 8.35e-02. acc. prob=0.78]
warmup: 40%|███▉ | 398/1000 [01:03<00:22, 26.50it/s, 127 steps of size 6.07e-02. acc. prob=0.78]
warmup: 40%|████ | 401/1000 [01:03<00:23, 25.31it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 40%|████ | 405/1000 [01:03<00:22, 26.80it/s, 63 steps of size 6.77e-02. acc. prob=0.78]
warmup: 41%|████ | 408/1000 [01:03<00:21, 27.11it/s, 31 steps of size 1.47e-01. acc. prob=0.78]
warmup: 41%|████ | 410/1000 [01:03<00:23, 24.69it/s, 127 steps of size 6.71e-02. acc. prob=0.78]
warmup: 41%|████▏ | 413/1000 [01:04<00:24, 23.97it/s, 63 steps of size 1.03e-01. acc. prob=0.78]
warmup: 42%|████▏ | 416/1000 [01:04<00:24, 23.64it/s, 63 steps of size 1.18e-01. acc. prob=0.78]
warmup: 42%|████▏ | 420/1000 [01:04<00:22, 25.71it/s, 63 steps of size 1.24e-01. acc. prob=0.78]
warmup: 42%|████▏ | 423/1000 [01:04<00:24, 23.58it/s, 127 steps of size 1.19e-01. acc. prob=0.78]
warmup: 43%|████▎ | 426/1000 [01:04<00:24, 23.25it/s, 127 steps of size 6.53e-02. acc. prob=0.78]
warmup: 43%|████▎ | 429/1000 [01:04<00:24, 23.00it/s, 63 steps of size 1.13e-01. acc. prob=0.78]
warmup: 43%|████▎ | 432/1000 [01:04<00:23, 24.00it/s, 63 steps of size 1.23e-01. acc. prob=0.78]
warmup: 44%|████▎ | 436/1000 [01:04<00:21, 25.89it/s, 63 steps of size 1.51e-01. acc. prob=0.79]
warmup: 44%|████▍ | 438/1000 [01:05<00:23, 23.52it/s, 127 steps of size 7.38e-02. acc. prob=0.78]
warmup: 44%|████▍ | 441/1000 [01:05<00:23, 23.35it/s, 63 steps of size 9.40e-02. acc. prob=0.78]
warmup: 44%|████▍ | 445/1000 [01:05<00:20, 26.89it/s, 31 steps of size 8.31e-02. acc. prob=0.78]
warmup: 45%|████▍ | 449/1000 [01:05<00:19, 27.95it/s, 63 steps of size 1.58e-01. acc. prob=0.79]
warmup: 45%|████▌ | 454/1000 [01:05<00:22, 24.72it/s, 255 steps of size 2.08e-02. acc. prob=0.78]
warmup: 46%|████▌ | 455/1000 [01:05<00:29, 18.53it/s, 255 steps of size 3.18e-02. acc. prob=0.78]
warmup: 46%|████▌ | 456/1000 [01:06<00:37, 14.49it/s, 255 steps of size 4.51e-02. acc. prob=0.78]
warmup: 46%|████▌ | 458/1000 [01:06<00:37, 14.62it/s, 63 steps of size 8.24e-02. acc. prob=0.78]
warmup: 46%|████▌ | 461/1000 [01:06<01:02, 8.67it/s, 767 steps of size 2.11e-02. acc. prob=0.78]
warmup: 46%|████▌ | 462/1000 [01:06<01:06, 8.05it/s, 255 steps of size 3.99e-02. acc. prob=0.78]
warmup: 46%|████▋ | 463/1000 [01:07<01:11, 7.49it/s, 255 steps of size 7.41e-02. acc. prob=0.78]
warmup: 46%|████▋ | 465/1000 [01:07<01:05, 8.13it/s, 255 steps of size 4.67e-02. acc. prob=0.78]
warmup: 47%|████▋ | 467/1000 [01:07<00:56, 9.51it/s, 63 steps of size 9.79e-02. acc. prob=0.78]
warmup: 47%|████▋ | 470/1000 [01:07<00:43, 12.31it/s, 127 steps of size 4.44e-02. acc. prob=0.78]
warmup: 47%|████▋ | 472/1000 [01:07<00:40, 12.92it/s, 63 steps of size 5.80e-02. acc. prob=0.78]
warmup: 47%|████▋ | 474/1000 [01:07<00:37, 14.05it/s, 31 steps of size 3.25e-02. acc. prob=0.78]
warmup: 48%|████▊ | 476/1000 [01:07<00:36, 14.33it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 48%|████▊ | 479/1000 [01:08<00:37, 13.98it/s, 255 steps of size 3.19e-02. acc. prob=0.78]
warmup: 48%|████▊ | 481/1000 [01:08<00:36, 14.25it/s, 63 steps of size 9.34e-02. acc. prob=0.78]
warmup: 48%|████▊ | 483/1000 [01:08<00:34, 15.12it/s, 127 steps of size 6.94e-02. acc. prob=0.78]
warmup: 49%|████▊ | 486/1000 [01:08<00:28, 18.00it/s, 31 steps of size 3.38e-02. acc. prob=0.78]
warmup: 49%|████▉ | 488/1000 [01:08<00:32, 15.52it/s, 127 steps of size 9.63e-02. acc. prob=0.78]
warmup: 49%|████▉ | 491/1000 [01:09<00:37, 13.64it/s, 255 steps of size 5.10e-02. acc. prob=0.78]
warmup: 49%|████▉ | 493/1000 [01:09<00:36, 13.94it/s, 127 steps of size 9.56e-02. acc. prob=0.78]
warmup: 50%|████▉ | 496/1000 [01:09<00:29, 16.82it/s, 31 steps of size 2.29e-02. acc. prob=0.78]
warmup: 50%|████▉ | 498/1000 [01:09<00:33, 15.00it/s, 127 steps of size 5.77e-02. acc. prob=0.78]
warmup: 50%|█████ | 500/1000 [01:09<00:31, 15.65it/s, 31 steps of size 6.30e-02. acc. prob=0.78]
sample: 50%|█████ | 502/1000 [01:09<00:32, 15.41it/s, 127 steps of size 6.30e-02. acc. prob=0.98]
sample: 50%|█████ | 505/1000 [01:09<00:28, 17.27it/s, 63 steps of size 6.30e-02. acc. prob=0.96]
sample: 51%|█████ | 508/1000 [01:09<00:26, 18.81it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 51%|█████ | 510/1000 [01:10<00:27, 17.69it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 51%|█████▏ | 513/1000 [01:10<00:28, 17.35it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 52%|█████▏ | 516/1000 [01:10<00:30, 15.64it/s, 191 steps of size 6.30e-02. acc. prob=0.95]
sample: 52%|█████▏ | 518/1000 [01:10<00:31, 15.13it/s, 63 steps of size 6.30e-02. acc. prob=0.95]
sample: 52%|█████▏ | 520/1000 [01:10<00:32, 14.83it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 52%|█████▏ | 521/1000 [01:10<00:34, 13.80it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 52%|█████▏ | 523/1000 [01:10<00:31, 15.12it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 53%|█████▎ | 526/1000 [01:11<00:28, 16.77it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 53%|█████▎ | 527/1000 [01:11<00:31, 15.12it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 53%|█████▎ | 530/1000 [01:11<00:27, 16.87it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 53%|█████▎ | 533/1000 [01:11<00:25, 18.18it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 54%|█████▎ | 536/1000 [01:11<00:24, 19.10it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 54%|█████▍ | 539/1000 [01:11<00:23, 19.84it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 54%|█████▍ | 542/1000 [01:11<00:24, 18.63it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 54%|█████▍ | 544/1000 [01:12<00:25, 17.55it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 55%|█████▍ | 547/1000 [01:12<00:24, 18.84it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 55%|█████▌ | 550/1000 [01:12<00:25, 17.94it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 55%|█████▌ | 552/1000 [01:12<00:26, 17.07it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 55%|█████▌ | 554/1000 [01:12<00:29, 15.11it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 56%|█████▌ | 557/1000 [01:12<00:28, 15.64it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 56%|█████▌ | 560/1000 [01:13<00:25, 17.35it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 56%|█████▌ | 562/1000 [01:13<00:26, 16.69it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 56%|█████▋ | 565/1000 [01:13<00:23, 18.34it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 57%|█████▋ | 567/1000 [01:13<00:24, 17.38it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 57%|█████▋ | 570/1000 [01:13<00:25, 17.20it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 57%|█████▋ | 573/1000 [01:13<00:22, 18.64it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 57%|█████▊ | 575/1000 [01:13<00:23, 18.45it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 58%|█████▊ | 577/1000 [01:14<00:24, 17.46it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 58%|█████▊ | 580/1000 [01:14<00:24, 17.29it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 58%|█████▊ | 582/1000 [01:14<00:25, 16.65it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 58%|█████▊ | 585/1000 [01:14<00:22, 18.40it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 59%|█████▊ | 587/1000 [01:14<00:23, 17.47it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 59%|█████▉ | 589/1000 [01:14<00:24, 16.73it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 59%|█████▉ | 591/1000 [01:14<00:25, 16.22it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 59%|█████▉ | 593/1000 [01:15<00:28, 14.48it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 60%|█████▉ | 596/1000 [01:15<00:24, 16.72it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 60%|█████▉ | 599/1000 [01:15<00:21, 18.32it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 60%|██████ | 602/1000 [01:15<00:20, 19.52it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 60%|██████ | 604/1000 [01:15<00:21, 18.21it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 61%|██████ | 606/1000 [01:15<00:22, 17.19it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 61%|██████ | 608/1000 [01:15<00:24, 16.19it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 61%|██████ | 610/1000 [01:16<00:26, 14.46it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 61%|██████▏ | 613/1000 [01:16<00:23, 16.34it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 62%|██████▏ | 615/1000 [01:16<00:24, 15.56it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 62%|██████▏ | 618/1000 [01:16<00:24, 15.67it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 62%|██████▏ | 621/1000 [01:16<00:22, 17.08it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 62%|██████▏ | 624/1000 [01:16<00:20, 17.99it/s, 63 steps of size 6.30e-02. acc. prob=0.93]
sample: 63%|██████▎ | 626/1000 [01:16<00:21, 17.07it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 63%|██████▎ | 628/1000 [01:17<00:24, 14.99it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 63%|██████▎ | 630/1000 [01:17<00:27, 13.66it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 63%|██████▎ | 632/1000 [01:17<00:26, 13.97it/s, 127 steps of size 6.30e-02. acc. prob=0.93]
sample: 64%|██████▎ | 635/1000 [01:17<00:22, 16.22it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 64%|██████▍ | 638/1000 [01:17<00:22, 16.41it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 64%|██████▍ | 640/1000 [01:17<00:22, 16.03it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 64%|██████▍ | 643/1000 [01:18<00:21, 16.28it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 64%|██████▍ | 645/1000 [01:18<00:26, 13.58it/s, 191 steps of size 6.30e-02. acc. prob=0.94]
sample: 65%|██████▍ | 648/1000 [01:18<00:24, 14.58it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 65%|██████▌ | 651/1000 [01:18<00:21, 16.42it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 65%|██████▌ | 654/1000 [01:18<00:20, 16.49it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 66%|██████▌ | 656/1000 [01:18<00:21, 16.02it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 66%|██████▌ | 659/1000 [01:19<00:19, 17.79it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 66%|██████▌ | 662/1000 [01:19<00:17, 19.10it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 66%|██████▋ | 664/1000 [01:19<00:18, 17.93it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 67%|██████▋ | 666/1000 [01:19<00:19, 17.14it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 67%|██████▋ | 669/1000 [01:19<00:17, 18.54it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 67%|██████▋ | 671/1000 [01:19<00:20, 15.77it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 67%|██████▋ | 674/1000 [01:19<00:18, 17.53it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 68%|██████▊ | 676/1000 [01:20<00:19, 16.67it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 68%|██████▊ | 679/1000 [01:20<00:17, 18.28it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 68%|██████▊ | 681/1000 [01:20<00:18, 17.42it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 68%|██████▊ | 684/1000 [01:20<00:16, 18.99it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 69%|██████▊ | 686/1000 [01:20<00:17, 17.83it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 69%|██████▉ | 689/1000 [01:20<00:17, 17.45it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 69%|██████▉ | 691/1000 [01:20<00:20, 15.34it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 69%|██████▉ | 694/1000 [01:21<00:17, 17.24it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 70%|██████▉ | 697/1000 [01:21<00:16, 18.62it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 70%|███████ | 700/1000 [01:21<00:15, 19.53it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 70%|███████ | 702/1000 [01:21<00:16, 18.11it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 70%|███████ | 704/1000 [01:21<00:17, 17.17it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 71%|███████ | 706/1000 [01:21<00:17, 16.54it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 71%|███████ | 709/1000 [01:21<00:15, 18.28it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 71%|███████ | 711/1000 [01:22<00:16, 17.16it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 71%|███████▏ | 713/1000 [01:22<00:19, 14.88it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 72%|███████▏ | 715/1000 [01:22<00:19, 14.78it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 72%|███████▏ | 718/1000 [01:22<00:18, 15.34it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 72%|███████▏ | 721/1000 [01:22<00:15, 17.92it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 72%|███████▏ | 724/1000 [01:22<00:14, 19.15it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 73%|███████▎ | 726/1000 [01:22<00:15, 17.84it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 73%|███████▎ | 729/1000 [01:23<00:15, 17.32it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 73%|███████▎ | 732/1000 [01:23<00:14, 18.53it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 74%|███████▎ | 735/1000 [01:23<00:13, 19.58it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 74%|███████▍ | 738/1000 [01:23<00:12, 20.56it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 74%|███████▍ | 740/1000 [01:23<00:13, 18.88it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 74%|███████▍ | 743/1000 [01:23<00:12, 19.79it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 75%|███████▍ | 746/1000 [01:23<00:11, 21.45it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 75%|███████▍ | 748/1000 [01:24<00:13, 19.37it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 75%|███████▌ | 750/1000 [01:24<00:15, 16.35it/s, 191 steps of size 6.30e-02. acc. prob=0.94]
sample: 75%|███████▌ | 752/1000 [01:24<00:16, 14.62it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 76%|███████▌ | 755/1000 [01:24<00:16, 15.26it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 76%|███████▌ | 758/1000 [01:24<00:15, 15.75it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 76%|███████▌ | 760/1000 [01:24<00:18, 13.33it/s, 191 steps of size 6.30e-02. acc. prob=0.94]
sample: 76%|███████▌ | 762/1000 [01:25<00:17, 13.60it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 76%|███████▋ | 763/1000 [01:25<00:19, 12.36it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 77%|███████▋ | 766/1000 [01:25<00:15, 14.81it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 77%|███████▋ | 769/1000 [01:25<00:13, 16.87it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 77%|███████▋ | 772/1000 [01:25<00:12, 18.31it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 78%|███████▊ | 775/1000 [01:25<00:12, 17.78it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 78%|███████▊ | 778/1000 [01:25<00:11, 19.05it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 78%|███████▊ | 781/1000 [01:26<00:10, 19.92it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 78%|███████▊ | 784/1000 [01:26<00:10, 20.50it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 79%|███████▊ | 787/1000 [01:26<00:11, 19.13it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 79%|███████▉ | 789/1000 [01:26<00:11, 18.00it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 79%|███████▉ | 791/1000 [01:26<00:13, 15.67it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 79%|███████▉ | 794/1000 [01:26<00:12, 16.10it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 80%|███████▉ | 797/1000 [01:27<00:12, 16.42it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 80%|███████▉ | 799/1000 [01:27<00:12, 16.12it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 80%|████████ | 801/1000 [01:27<00:12, 15.80it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 80%|████████ | 804/1000 [01:27<00:13, 14.95it/s, 191 steps of size 6.30e-02. acc. prob=0.94]
sample: 81%|████████ | 806/1000 [01:27<00:14, 13.85it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 81%|████████ | 809/1000 [01:27<00:11, 15.97it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 81%|████████ | 811/1000 [01:27<00:11, 16.40it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 81%|████████▏ | 814/1000 [01:28<00:10, 18.14it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 82%|████████▏ | 817/1000 [01:28<00:08, 20.42it/s, 31 steps of size 6.30e-02. acc. prob=0.94]
sample: 82%|████████▏ | 820/1000 [01:28<00:08, 21.07it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 82%|████████▏ | 823/1000 [01:28<00:09, 19.52it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 82%|████████▎ | 825/1000 [01:28<00:09, 18.27it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 83%|████████▎ | 828/1000 [01:28<00:08, 19.47it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 83%|████████▎ | 831/1000 [01:28<00:08, 20.36it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 83%|████████▎ | 834/1000 [01:29<00:07, 22.18it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 84%|████████▎ | 836/1000 [01:29<00:09, 18.19it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 84%|████████▍ | 838/1000 [01:29<00:10, 15.81it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 84%|████████▍ | 840/1000 [01:29<00:11, 14.24it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 84%|████████▍ | 843/1000 [01:29<00:09, 16.40it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 85%|████████▍ | 846/1000 [01:29<00:08, 18.16it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 85%|████████▍ | 849/1000 [01:30<00:08, 17.72it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 85%|████████▌ | 851/1000 [01:30<00:08, 17.09it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 85%|████████▌ | 853/1000 [01:30<00:08, 16.45it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 86%|████████▌ | 855/1000 [01:30<00:09, 16.05it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 86%|████████▌ | 858/1000 [01:30<00:08, 16.35it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 86%|████████▌ | 861/1000 [01:30<00:07, 18.12it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 86%|████████▋ | 864/1000 [01:30<00:07, 19.28it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 87%|████████▋ | 867/1000 [01:30<00:06, 20.31it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 87%|████████▋ | 870/1000 [01:31<00:06, 19.01it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 87%|████████▋ | 872/1000 [01:31<00:07, 17.84it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 87%|████████▋ | 874/1000 [01:31<00:07, 17.00it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 88%|████████▊ | 877/1000 [01:31<00:06, 18.54it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 88%|████████▊ | 880/1000 [01:31<00:06, 18.11it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 88%|████████▊ | 883/1000 [01:31<00:06, 17.73it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 89%|████████▊ | 886/1000 [01:32<00:05, 19.00it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 89%|████████▉ | 888/1000 [01:32<00:06, 17.96it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 89%|████████▉ | 890/1000 [01:32<00:06, 17.11it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 89%|████████▉ | 893/1000 [01:32<00:05, 18.66it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 90%|████████▉ | 895/1000 [01:32<00:06, 16.00it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 90%|████████▉ | 897/1000 [01:32<00:06, 15.72it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 90%|████████▉ | 899/1000 [01:32<00:06, 15.48it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 90%|█████████ | 902/1000 [01:33<00:06, 16.00it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 90%|█████████ | 905/1000 [01:33<00:05, 17.79it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 91%|█████████ | 907/1000 [01:33<00:05, 15.51it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 91%|█████████ | 909/1000 [01:33<00:05, 15.33it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 91%|█████████ | 912/1000 [01:33<00:05, 15.91it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 92%|█████████▏| 915/1000 [01:33<00:05, 16.26it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 92%|█████████▏| 918/1000 [01:33<00:04, 17.81it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 92%|█████████▏| 921/1000 [01:34<00:04, 19.07it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 92%|█████████▏| 924/1000 [01:34<00:04, 18.29it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 93%|█████████▎| 926/1000 [01:34<00:04, 15.99it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 93%|█████████▎| 928/1000 [01:34<00:04, 15.78it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 93%|█████████▎| 931/1000 [01:34<00:03, 17.63it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 93%|█████████▎| 934/1000 [01:34<00:03, 18.93it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 94%|█████████▎| 936/1000 [01:35<00:03, 17.87it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 94%|█████████▍| 939/1000 [01:35<00:03, 17.46it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 94%|█████████▍| 941/1000 [01:35<00:03, 16.56it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 94%|█████████▍| 943/1000 [01:35<00:03, 15.95it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 95%|█████████▍| 946/1000 [01:35<00:03, 17.84it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 95%|█████████▍| 948/1000 [01:35<00:03, 16.82it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 95%|█████████▌| 950/1000 [01:35<00:03, 14.78it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 95%|█████████▌| 953/1000 [01:36<00:02, 16.83it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 96%|█████████▌| 956/1000 [01:36<00:02, 18.37it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 96%|█████████▌| 959/1000 [01:36<00:02, 17.78it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 96%|█████████▌| 961/1000 [01:36<00:02, 16.96it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 96%|█████████▋| 963/1000 [01:36<00:02, 17.09it/s, 31 steps of size 6.30e-02. acc. prob=0.94]
sample: 96%|█████████▋| 965/1000 [01:36<00:02, 15.01it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 97%|█████████▋| 967/1000 [01:36<00:02, 15.09it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 97%|█████████▋| 969/1000 [01:37<00:02, 13.73it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 97%|█████████▋| 972/1000 [01:37<00:01, 16.18it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 97%|█████████▋| 974/1000 [01:37<00:01, 15.86it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 98%|█████████▊| 977/1000 [01:37<00:01, 17.76it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 98%|█████████▊| 980/1000 [01:37<00:01, 17.42it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 98%|█████████▊| 982/1000 [01:37<00:01, 16.71it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 98%|█████████▊| 985/1000 [01:37<00:00, 18.32it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 99%|█████████▉| 988/1000 [01:38<00:00, 19.65it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 99%|█████████▉| 990/1000 [01:38<00:00, 18.22it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 99%|█████████▉| 993/1000 [01:38<00:00, 20.47it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 100%|█████████▉| 995/1000 [01:38<00:00, 18.76it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 100%|█████████▉| 998/1000 [01:38<00:00, 19.82it/s, 63 steps of size 6.30e-02. acc. prob=0.94]
sample: 100%|██████████| 1000/1000 [01:38<00:00, 18.28it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
sample: 100%|██████████| 1000/1000 [01:38<00:00, 10.13it/s, 127 steps of size 6.30e-02. acc. prob=0.94]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 0%| | 0/1000 [00:11<?, ?it/s, 511 steps of size 1.50e-03. acc. prob=0.71]
warmup: 4%|▍ | 42/1000 [00:11<04:31, 3.53it/s, 1023 steps of size 2.57e-03. acc. prob=0.72]
warmup: 4%|▍ | 43/1000 [00:12<04:46, 3.34it/s, 1023 steps of size 4.43e-03. acc. prob=0.73]
warmup: 4%|▍ | 44/1000 [00:13<05:06, 3.12it/s, 1023 steps of size 5.88e-03. acc. prob=0.73]
warmup: 4%|▍ | 45/1000 [00:13<05:08, 3.09it/s, 511 steps of size 8.99e-03. acc. prob=0.74]
warmup: 5%|▍ | 47/1000 [00:14<05:13, 3.04it/s, 1023 steps of size 2.31e-03. acc. prob=0.73]
warmup: 5%|▍ | 48/1000 [00:15<05:47, 2.74it/s, 1023 steps of size 3.90e-03. acc. prob=0.73]
warmup: 5%|▍ | 49/1000 [00:15<05:46, 2.75it/s, 511 steps of size 4.73e-03. acc. prob=0.73]
warmup: 5%|▌ | 50/1000 [00:15<05:43, 2.76it/s, 511 steps of size 4.33e-03. acc. prob=0.73]
warmup: 5%|▌ | 51/1000 [00:16<06:39, 2.38it/s, 1023 steps of size 6.44e-03. acc. prob=0.74]
warmup: 5%|▌ | 52/1000 [00:17<06:26, 2.45it/s, 511 steps of size 7.53e-03. acc. prob=0.74]
warmup: 5%|▌ | 53/1000 [00:17<06:15, 2.52it/s, 511 steps of size 6.73e-03. acc. prob=0.74]
warmup: 5%|▌ | 54/1000 [00:17<05:25, 2.91it/s, 255 steps of size 3.08e-03. acc. prob=0.73]
warmup: 6%|▌ | 55/1000 [00:18<06:50, 2.30it/s, 1023 steps of size 4.59e-03. acc. prob=0.74]
warmup: 6%|▌ | 56/1000 [00:18<07:58, 1.97it/s, 1023 steps of size 6.92e-03. acc. prob=0.74]
warmup: 6%|▌ | 57/1000 [00:19<06:32, 2.40it/s, 255 steps of size 4.95e-03. acc. prob=0.74]
warmup: 6%|▌ | 58/1000 [00:19<06:13, 2.52it/s, 511 steps of size 7.40e-03. acc. prob=0.74]
warmup: 6%|▌ | 59/1000 [00:19<06:01, 2.60it/s, 511 steps of size 1.17e-02. acc. prob=0.75]
warmup: 6%|▌ | 61/1000 [00:20<05:48, 2.70it/s, 1023 steps of size 3.33e-03. acc. prob=0.74]
warmup: 6%|▌ | 62/1000 [00:21<07:03, 2.21it/s, 1023 steps of size 4.42e-03. acc. prob=0.74]
warmup: 6%|▋ | 63/1000 [00:21<06:39, 2.34it/s, 511 steps of size 6.95e-03. acc. prob=0.75]
warmup: 6%|▋ | 64/1000 [00:21<06:20, 2.46it/s, 511 steps of size 4.89e-03. acc. prob=0.74]
warmup: 6%|▋ | 65/1000 [00:22<06:06, 2.55it/s, 511 steps of size 6.47e-03. acc. prob=0.75]
warmup: 7%|▋ | 66/1000 [00:22<05:11, 3.00it/s, 255 steps of size 6.88e-03. acc. prob=0.75]
warmup: 7%|▋ | 67/1000 [00:22<05:18, 2.93it/s, 511 steps of size 7.51e-03. acc. prob=0.75]
warmup: 7%|▋ | 68/1000 [00:23<05:22, 2.89it/s, 511 steps of size 2.41e-03. acc. prob=0.74]
warmup: 7%|▋ | 69/1000 [00:23<07:02, 2.21it/s, 1023 steps of size 3.84e-03. acc. prob=0.75]
warmup: 7%|▋ | 70/1000 [00:24<08:12, 1.89it/s, 1023 steps of size 6.03e-03. acc. prob=0.75]
warmup: 7%|▋ | 71/1000 [00:24<07:22, 2.10it/s, 511 steps of size 8.96e-03. acc. prob=0.75]
warmup: 7%|▋ | 73/1000 [00:25<06:34, 2.35it/s, 1023 steps of size 3.04e-03. acc. prob=0.75]
warmup: 7%|▋ | 74/1000 [00:26<07:44, 1.99it/s, 1023 steps of size 3.58e-03. acc. prob=0.75]
warmup: 8%|▊ | 75/1000 [00:27<08:36, 1.79it/s, 1023 steps of size 5.59e-03. acc. prob=0.75]
warmup: 8%|▊ | 76/1000 [00:27<07:44, 1.99it/s, 511 steps of size 5.60e-03. acc. prob=0.75]
warmup: 8%|▊ | 77/1000 [00:27<07:06, 2.17it/s, 511 steps of size 8.70e-03. acc. prob=0.75]
warmup: 8%|▊ | 78/1000 [00:28<06:37, 2.32it/s, 511 steps of size 3.44e-03. acc. prob=0.75]
warmup: 8%|▊ | 79/1000 [00:28<07:53, 1.94it/s, 1023 steps of size 4.80e-03. acc. prob=0.75]
warmup: 8%|▊ | 80/1000 [00:29<08:57, 1.71it/s, 1023 steps of size 7.37e-03. acc. prob=0.75]
warmup: 8%|▊ | 81/1000 [00:30<07:57, 1.92it/s, 511 steps of size 8.81e-03. acc. prob=0.76]
warmup: 8%|▊ | 83/1000 [00:30<07:04, 2.16it/s, 1023 steps of size 3.07e-03. acc. prob=0.75]
warmup: 8%|▊ | 84/1000 [00:31<08:00, 1.91it/s, 1023 steps of size 4.24e-03. acc. prob=0.75]
warmup: 8%|▊ | 85/1000 [00:32<08:43, 1.75it/s, 1023 steps of size 6.28e-03. acc. prob=0.76]
warmup: 9%|▊ | 86/1000 [00:32<07:51, 1.94it/s, 511 steps of size 9.06e-03. acc. prob=0.76]
warmup: 9%|▉ | 88/1000 [00:33<06:51, 2.22it/s, 1023 steps of size 3.04e-03. acc. prob=0.75]
warmup: 9%|▉ | 89/1000 [00:34<07:46, 1.95it/s, 1023 steps of size 4.09e-03. acc. prob=0.75]
warmup: 9%|▉ | 90/1000 [00:34<08:32, 1.78it/s, 1023 steps of size 5.08e-03. acc. prob=0.76]
warmup: 9%|▉ | 91/1000 [00:35<09:06, 1.66it/s, 1023 steps of size 6.85e-03. acc. prob=0.76]
warmup: 9%|▉ | 92/1000 [00:35<08:04, 1.87it/s, 511 steps of size 5.56e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:36<07:18, 2.07it/s, 511 steps of size 8.33e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:36<05:54, 2.55it/s, 231 steps of size 3.26e-03. acc. prob=0.75]
warmup: 10%|▉ | 95/1000 [00:36<05:44, 2.62it/s, 511 steps of size 3.80e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:37<07:11, 2.10it/s, 1023 steps of size 3.89e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:38<08:12, 1.83it/s, 1023 steps of size 5.35e-03. acc. prob=0.76]
warmup: 10%|▉ | 98/1000 [00:38<08:55, 1.68it/s, 1023 steps of size 6.54e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:39<07:50, 1.91it/s, 511 steps of size 7.32e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:39<07:04, 2.12it/s, 511 steps of size 7.14e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:39<06:32, 2.29it/s, 511 steps of size 1.02e-01. acc. prob=0.76]
warmup: 10%|█ | 104/1000 [00:40<03:31, 4.23it/s, 255 steps of size 5.14e-02. acc. prob=0.76]
warmup: 11%|█ | 106/1000 [00:40<02:35, 5.77it/s, 31 steps of size 1.53e-01. acc. prob=0.76]
warmup: 11%|█ | 108/1000 [00:40<02:10, 6.83it/s, 255 steps of size 2.39e-02. acc. prob=0.76]
warmup: 11%|█ | 110/1000 [00:40<01:53, 7.83it/s, 127 steps of size 8.14e-02. acc. prob=0.76]
warmup: 11%|█▏ | 113/1000 [00:40<01:20, 11.03it/s, 31 steps of size 1.49e-01. acc. prob=0.77]
warmup: 12%|█▏ | 115/1000 [00:41<01:45, 8.37it/s, 511 steps of size 3.07e-02. acc. prob=0.76]
warmup: 12%|█▏ | 117/1000 [00:41<01:32, 9.56it/s, 63 steps of size 1.05e-01. acc. prob=0.77]
warmup: 12%|█▏ | 120/1000 [00:41<01:25, 10.32it/s, 255 steps of size 3.41e-02. acc. prob=0.76]
warmup: 12%|█▏ | 122/1000 [00:41<01:17, 11.27it/s, 63 steps of size 8.09e-02. acc. prob=0.77]
warmup: 12%|█▏ | 124/1000 [00:41<01:15, 11.68it/s, 95 steps of size 1.37e-01. acc. prob=0.77]
warmup: 13%|█▎ | 127/1000 [00:41<01:01, 14.19it/s, 127 steps of size 4.60e-02. acc. prob=0.77]
warmup: 13%|█▎ | 129/1000 [00:42<01:00, 14.32it/s, 63 steps of size 9.97e-02. acc. prob=0.77]
warmup: 13%|█▎ | 131/1000 [00:42<01:00, 14.44it/s, 127 steps of size 9.69e-02. acc. prob=0.77]
warmup: 14%|█▎ | 135/1000 [00:42<00:59, 14.54it/s, 255 steps of size 3.26e-02. acc. prob=0.77]
warmup: 14%|█▎ | 137/1000 [00:42<00:59, 14.57it/s, 63 steps of size 5.64e-02. acc. prob=0.77]
warmup: 14%|█▍ | 139/1000 [00:42<00:58, 14.60it/s, 63 steps of size 1.19e-01. acc. prob=0.77]
warmup: 14%|█▍ | 141/1000 [00:42<00:56, 15.33it/s, 127 steps of size 7.79e-02. acc. prob=0.77]
warmup: 14%|█▍ | 145/1000 [00:43<00:56, 15.11it/s, 255 steps of size 6.14e-02. acc. prob=0.77]
warmup: 15%|█▍ | 148/1000 [00:43<00:50, 16.90it/s, 63 steps of size 1.11e-01. acc. prob=0.77]
warmup: 15%|█▌ | 154/1000 [00:43<00:56, 15.01it/s, 511 steps of size 2.21e-02. acc. prob=0.77]
warmup: 16%|█▌ | 155/1000 [00:43<01:05, 12.82it/s, 255 steps of size 3.34e-02. acc. prob=0.77]
warmup: 16%|█▌ | 157/1000 [00:44<01:03, 13.29it/s, 63 steps of size 6.00e-02. acc. prob=0.77]
warmup: 16%|█▌ | 161/1000 [00:44<00:58, 14.36it/s, 255 steps of size 7.18e-02. acc. prob=0.77]
warmup: 16%|█▋ | 163/1000 [00:44<00:57, 14.64it/s, 63 steps of size 1.68e-01. acc. prob=0.77]
warmup: 16%|█▋ | 165/1000 [00:44<01:03, 13.20it/s, 255 steps of size 3.36e-02. acc. prob=0.77]
warmup: 17%|█▋ | 167/1000 [00:44<01:01, 13.64it/s, 63 steps of size 1.14e-01. acc. prob=0.77]
warmup: 17%|█▋ | 171/1000 [00:44<00:56, 14.69it/s, 255 steps of size 3.76e-02. acc. prob=0.77]
warmup: 17%|█▋ | 173/1000 [00:45<01:00, 13.73it/s, 127 steps of size 1.03e-01. acc. prob=0.77]
warmup: 18%|█▊ | 177/1000 [00:45<00:57, 14.22it/s, 255 steps of size 2.90e-02. acc. prob=0.77]
warmup: 18%|█▊ | 178/1000 [00:45<01:08, 11.94it/s, 255 steps of size 4.97e-02. acc. prob=0.77]
warmup: 18%|█▊ | 180/1000 [00:45<01:04, 12.66it/s, 63 steps of size 9.98e-02. acc. prob=0.77]
warmup: 18%|█▊ | 183/1000 [00:45<01:04, 12.67it/s, 255 steps of size 4.43e-02. acc. prob=0.77]
warmup: 18%|█▊ | 185/1000 [00:46<01:01, 13.20it/s, 63 steps of size 1.21e-01. acc. prob=0.77]
warmup: 19%|█▊ | 187/1000 [00:46<00:59, 13.62it/s, 127 steps of size 5.60e-02. acc. prob=0.77]
warmup: 19%|█▉ | 190/1000 [00:46<00:48, 16.61it/s, 31 steps of size 1.20e-01. acc. prob=0.78]
warmup: 19%|█▉ | 193/1000 [00:46<00:42, 19.14it/s, 31 steps of size 3.82e-02. acc. prob=0.77]
warmup: 20%|█▉ | 195/1000 [00:46<00:49, 16.42it/s, 127 steps of size 8.47e-02. acc. prob=0.78]
warmup: 20%|█▉ | 198/1000 [00:46<00:48, 16.60it/s, 127 steps of size 9.22e-02. acc. prob=0.78]
warmup: 20%|██ | 201/1000 [00:46<00:47, 16.69it/s, 127 steps of size 7.82e-02. acc. prob=0.78]
warmup: 20%|██ | 203/1000 [00:47<00:47, 16.91it/s, 95 steps of size 1.16e-01. acc. prob=0.78]
warmup: 21%|██ | 206/1000 [00:47<00:40, 19.83it/s, 13 steps of size 4.79e-02. acc. prob=0.77]
warmup: 21%|██ | 208/1000 [00:47<00:47, 16.68it/s, 127 steps of size 1.15e-01. acc. prob=0.78]
warmup: 21%|██ | 210/1000 [00:47<00:46, 16.99it/s, 127 steps of size 8.39e-02. acc. prob=0.78]
warmup: 21%|██▏ | 213/1000 [00:47<00:40, 19.62it/s, 31 steps of size 7.56e-02. acc. prob=0.78]
warmup: 22%|██▏ | 216/1000 [00:47<00:38, 20.53it/s, 63 steps of size 1.11e-01. acc. prob=0.78]
warmup: 22%|██▏ | 219/1000 [00:47<00:37, 21.09it/s, 127 steps of size 6.51e-02. acc. prob=0.78]
warmup: 22%|██▏ | 222/1000 [00:47<00:36, 21.52it/s, 63 steps of size 1.63e-01. acc. prob=0.78]
warmup: 22%|██▏ | 224/1000 [00:48<00:37, 20.60it/s, 127 steps of size 5.11e-02. acc. prob=0.78]
warmup: 23%|██▎ | 226/1000 [00:48<00:41, 18.82it/s, 63 steps of size 1.09e-01. acc. prob=0.78]
warmup: 23%|██▎ | 229/1000 [00:48<00:38, 19.97it/s, 63 steps of size 1.40e-01. acc. prob=0.78]
warmup: 23%|██▎ | 231/1000 [00:48<00:39, 19.36it/s, 127 steps of size 6.67e-02. acc. prob=0.78]
warmup: 23%|██▎ | 233/1000 [00:48<00:47, 16.25it/s, 127 steps of size 1.32e-01. acc. prob=0.78]
warmup: 24%|██▎ | 236/1000 [00:48<00:44, 17.14it/s, 127 steps of size 5.23e-02. acc. prob=0.78]
warmup: 24%|██▍ | 238/1000 [00:48<00:46, 16.50it/s, 63 steps of size 1.15e-01. acc. prob=0.78]
warmup: 24%|██▍ | 240/1000 [00:49<00:47, 16.10it/s, 127 steps of size 1.00e-01. acc. prob=0.78]
warmup: 24%|██▍ | 243/1000 [00:49<00:42, 18.01it/s, 63 steps of size 1.14e-01. acc. prob=0.78]
warmup: 25%|██▍ | 246/1000 [00:49<00:37, 20.33it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 25%|██▍ | 249/1000 [00:49<00:33, 22.12it/s, 31 steps of size 7.90e-02. acc. prob=0.78]
warmup: 25%|██▌ | 254/1000 [00:49<00:36, 20.21it/s, 255 steps of size 1.82e-02. acc. prob=0.77]
warmup: 26%|██▌ | 255/1000 [00:49<00:47, 15.62it/s, 255 steps of size 2.77e-02. acc. prob=0.78]
warmup: 26%|██▌ | 256/1000 [00:50<00:59, 12.56it/s, 255 steps of size 4.59e-02. acc. prob=0.78]
warmup: 26%|██▌ | 258/1000 [00:50<00:56, 13.17it/s, 63 steps of size 1.09e-01. acc. prob=0.78]
warmup: 26%|██▌ | 261/1000 [00:50<00:55, 13.22it/s, 255 steps of size 3.56e-02. acc. prob=0.78]
warmup: 26%|██▋ | 263/1000 [00:50<00:53, 13.70it/s, 63 steps of size 1.09e-01. acc. prob=0.78]
warmup: 27%|██▋ | 266/1000 [00:50<00:44, 16.65it/s, 63 steps of size 7.18e-02. acc. prob=0.78]
warmup: 27%|██▋ | 269/1000 [00:50<00:49, 14.77it/s, 255 steps of size 4.06e-02. acc. prob=0.78]
warmup: 27%|██▋ | 271/1000 [00:51<00:49, 14.84it/s, 63 steps of size 5.83e-02. acc. prob=0.78]
warmup: 27%|██▋ | 274/1000 [00:51<00:41, 17.62it/s, 31 steps of size 1.23e-01. acc. prob=0.78]
warmup: 28%|██▊ | 276/1000 [00:51<00:56, 12.82it/s, 383 steps of size 5.08e-02. acc. prob=0.78]
warmup: 28%|██▊ | 278/1000 [00:51<00:54, 13.31it/s, 63 steps of size 1.52e-01. acc. prob=0.78]
warmup: 28%|██▊ | 280/1000 [00:51<00:56, 12.65it/s, 255 steps of size 3.12e-02. acc. prob=0.78]
warmup: 28%|██▊ | 282/1000 [00:51<00:54, 13.22it/s, 63 steps of size 8.49e-02. acc. prob=0.78]
warmup: 28%|██▊ | 285/1000 [00:52<00:47, 14.98it/s, 127 steps of size 3.35e-02. acc. prob=0.78]
warmup: 29%|██▊ | 286/1000 [00:52<00:59, 11.98it/s, 255 steps of size 5.60e-02. acc. prob=0.78]
warmup: 29%|██▉ | 289/1000 [00:52<00:46, 15.32it/s, 31 steps of size 9.74e-02. acc. prob=0.78]
warmup: 29%|██▉ | 292/1000 [00:52<00:42, 16.84it/s, 127 steps of size 5.87e-02. acc. prob=0.78]
warmup: 30%|██▉ | 295/1000 [00:52<00:38, 18.23it/s, 63 steps of size 1.25e-01. acc. prob=0.78]
warmup: 30%|██▉ | 297/1000 [00:52<00:39, 17.94it/s, 127 steps of size 5.74e-02. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:52<00:33, 20.63it/s, 16 steps of size 3.94e-02. acc. prob=0.78]
warmup: 30%|███ | 302/1000 [00:52<00:37, 18.73it/s, 63 steps of size 8.71e-02. acc. prob=0.78]
warmup: 30%|███ | 305/1000 [00:53<00:35, 19.78it/s, 63 steps of size 4.13e-02. acc. prob=0.78]
warmup: 31%|███ | 307/1000 [00:53<00:37, 18.30it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 31%|███ | 310/1000 [00:53<00:34, 20.11it/s, 127 steps of size 5.89e-02. acc. prob=0.78]
warmup: 31%|███▏ | 313/1000 [00:53<00:32, 20.86it/s, 63 steps of size 9.36e-02. acc. prob=0.78]
warmup: 32%|███▏ | 316/1000 [00:53<00:41, 16.43it/s, 255 steps of size 4.58e-02. acc. prob=0.78]
warmup: 32%|███▏ | 318/1000 [00:53<00:45, 14.86it/s, 127 steps of size 9.39e-02. acc. prob=0.78]
warmup: 32%|███▏ | 321/1000 [00:54<00:40, 16.77it/s, 63 steps of size 1.25e-01. acc. prob=0.78]
warmup: 32%|███▏ | 324/1000 [00:54<00:35, 19.16it/s, 31 steps of size 6.95e-02. acc. prob=0.78]
warmup: 33%|███▎ | 327/1000 [00:54<00:33, 20.13it/s, 63 steps of size 1.67e-01. acc. prob=0.78]
warmup: 33%|███▎ | 329/1000 [00:54<00:39, 16.97it/s, 255 steps of size 4.95e-02. acc. prob=0.78]
warmup: 33%|███▎ | 331/1000 [00:54<00:40, 16.47it/s, 63 steps of size 1.00e-01. acc. prob=0.78]
warmup: 33%|███▎ | 334/1000 [00:54<00:38, 17.35it/s, 127 steps of size 4.68e-02. acc. prob=0.78]
warmup: 34%|███▎ | 336/1000 [00:54<00:39, 16.73it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 34%|███▍ | 339/1000 [00:55<00:37, 17.58it/s, 127 steps of size 5.39e-02. acc. prob=0.78]
warmup: 34%|███▍ | 342/1000 [00:55<00:37, 17.37it/s, 127 steps of size 6.67e-02. acc. prob=0.78]
warmup: 34%|███▍ | 345/1000 [00:55<00:34, 18.87it/s, 63 steps of size 7.99e-02. acc. prob=0.78]
warmup: 35%|███▍ | 348/1000 [00:55<00:32, 19.87it/s, 63 steps of size 1.27e-01. acc. prob=0.78]
warmup: 35%|███▌ | 351/1000 [00:55<00:31, 20.45it/s, 127 steps of size 4.81e-02. acc. prob=0.78]
warmup: 35%|███▌ | 353/1000 [00:55<00:34, 18.85it/s, 63 steps of size 8.43e-02. acc. prob=0.78]
warmup: 36%|███▌ | 356/1000 [00:55<00:32, 19.92it/s, 63 steps of size 1.31e-01. acc. prob=0.78]
warmup: 36%|███▌ | 359/1000 [00:56<00:29, 21.80it/s, 63 steps of size 8.73e-02. acc. prob=0.78]
warmup: 36%|███▋ | 363/1000 [00:56<00:24, 25.62it/s, 31 steps of size 1.22e-01. acc. prob=0.78]
warmup: 37%|███▋ | 366/1000 [00:56<00:25, 24.68it/s, 63 steps of size 9.73e-02. acc. prob=0.78]
warmup: 37%|███▋ | 370/1000 [00:56<00:23, 26.37it/s, 63 steps of size 7.92e-02. acc. prob=0.78]
warmup: 37%|███▋ | 374/1000 [00:56<00:21, 28.53it/s, 63 steps of size 7.23e-02. acc. prob=0.78]
warmup: 38%|███▊ | 377/1000 [00:56<00:22, 28.14it/s, 31 steps of size 7.01e-02. acc. prob=0.78]
warmup: 38%|███▊ | 382/1000 [00:56<00:21, 29.37it/s, 127 steps of size 5.91e-02. acc. prob=0.78]
warmup: 38%|███▊ | 384/1000 [00:56<00:27, 22.71it/s, 127 steps of size 8.73e-02. acc. prob=0.78]
warmup: 39%|███▊ | 387/1000 [00:57<00:29, 20.76it/s, 127 steps of size 5.48e-02. acc. prob=0.78]
warmup: 39%|███▉ | 389/1000 [00:57<00:31, 19.19it/s, 63 steps of size 7.42e-02. acc. prob=0.78]
warmup: 39%|███▉ | 392/1000 [00:57<00:28, 21.13it/s, 31 steps of size 4.18e-02. acc. prob=0.78]
warmup: 39%|███▉ | 394/1000 [00:57<00:31, 19.32it/s, 63 steps of size 7.12e-02. acc. prob=0.78]
warmup: 40%|███▉ | 397/1000 [00:57<00:29, 20.29it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 40%|████ | 400/1000 [00:57<00:27, 22.08it/s, 63 steps of size 8.86e-02. acc. prob=0.78]
warmup: 40%|████ | 403/1000 [00:57<00:26, 22.29it/s, 63 steps of size 1.36e-01. acc. prob=0.78]
warmup: 41%|████ | 406/1000 [00:58<00:29, 20.17it/s, 127 steps of size 7.46e-02. acc. prob=0.78]
warmup: 41%|████ | 408/1000 [00:58<00:31, 18.75it/s, 127 steps of size 8.36e-02. acc. prob=0.78]
warmup: 41%|████ | 410/1000 [00:58<00:31, 18.58it/s, 31 steps of size 7.54e-02. acc. prob=0.78]
warmup: 41%|████▏ | 413/1000 [00:58<00:29, 19.75it/s, 63 steps of size 1.45e-01. acc. prob=0.78]
warmup: 42%|████▏ | 417/1000 [00:58<00:26, 22.34it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 42%|████▏ | 420/1000 [00:58<00:24, 23.64it/s, 63 steps of size 7.36e-02. acc. prob=0.78]
warmup: 42%|████▏ | 423/1000 [00:58<00:23, 24.60it/s, 31 steps of size 9.88e-02. acc. prob=0.78]
warmup: 43%|████▎ | 426/1000 [00:58<00:22, 25.20it/s, 63 steps of size 7.56e-02. acc. prob=0.78]
warmup: 43%|████▎ | 429/1000 [00:59<00:23, 24.49it/s, 63 steps of size 9.39e-02. acc. prob=0.78]
warmup: 43%|████▎ | 432/1000 [00:59<00:23, 23.91it/s, 63 steps of size 8.41e-02. acc. prob=0.78]
warmup: 44%|████▎ | 435/1000 [00:59<00:22, 24.79it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 44%|████▍ | 438/1000 [00:59<00:22, 25.42it/s, 63 steps of size 6.81e-02. acc. prob=0.78]
warmup: 44%|████▍ | 441/1000 [00:59<00:22, 24.50it/s, 63 steps of size 1.12e-01. acc. prob=0.78]
warmup: 44%|████▍ | 444/1000 [00:59<00:22, 25.21it/s, 63 steps of size 9.11e-02. acc. prob=0.78]
warmup: 45%|████▍ | 448/1000 [00:59<00:19, 28.30it/s, 31 steps of size 1.08e-01. acc. prob=0.78]
warmup: 45%|████▌ | 451/1000 [00:59<00:21, 25.03it/s, 127 steps of size 6.57e-01. acc. prob=0.78]
warmup: 45%|████▌ | 454/1000 [01:00<00:21, 25.50it/s, 127 steps of size 6.37e-02. acc. prob=0.78]
warmup: 46%|████▌ | 457/1000 [01:00<00:20, 26.00it/s, 31 steps of size 2.19e-02. acc. prob=0.78]
warmup: 46%|████▌ | 458/1000 [01:00<00:29, 18.10it/s, 255 steps of size 3.92e-02. acc. prob=0.78]
warmup: 46%|████▌ | 460/1000 [01:00<00:31, 17.16it/s, 63 steps of size 1.29e-01. acc. prob=0.78]
warmup: 46%|████▋ | 465/1000 [01:00<00:25, 21.09it/s, 127 steps of size 6.08e-02. acc. prob=0.78]
warmup: 47%|████▋ | 469/1000 [01:00<00:29, 18.23it/s, 255 steps of size 2.62e-02. acc. prob=0.78]
warmup: 47%|████▋ | 470/1000 [01:01<00:41, 12.73it/s, 383 steps of size 4.72e-02. acc. prob=0.78]
warmup: 47%|████▋ | 472/1000 [01:01<00:39, 13.27it/s, 63 steps of size 1.19e-01. acc. prob=0.78]
warmup: 48%|████▊ | 475/1000 [01:01<00:35, 14.88it/s, 127 steps of size 5.09e-02. acc. prob=0.78]
warmup: 48%|████▊ | 477/1000 [01:01<00:35, 14.93it/s, 63 steps of size 1.29e-01. acc. prob=0.78]
warmup: 48%|████▊ | 479/1000 [01:01<00:33, 15.59it/s, 127 steps of size 9.28e-02. acc. prob=0.78]
warmup: 48%|████▊ | 481/1000 [01:01<00:38, 13.58it/s, 255 steps of size 4.65e-02. acc. prob=0.78]
warmup: 48%|████▊ | 483/1000 [01:02<00:37, 13.97it/s, 63 steps of size 1.31e-01. acc. prob=0.78]
warmup: 48%|████▊ | 485/1000 [01:02<00:34, 15.03it/s, 127 steps of size 5.11e-02. acc. prob=0.78]
warmup: 49%|████▊ | 487/1000 [01:02<00:34, 15.03it/s, 63 steps of size 1.39e-01. acc. prob=0.78]
warmup: 49%|████▉ | 490/1000 [01:02<00:31, 16.35it/s, 127 steps of size 5.65e-02. acc. prob=0.78]
warmup: 49%|████▉ | 492/1000 [01:02<00:32, 15.85it/s, 63 steps of size 1.52e-01. acc. prob=0.78]
warmup: 50%|████▉ | 495/1000 [01:02<00:32, 15.42it/s, 255 steps of size 3.99e-02. acc. prob=0.78]
warmup: 50%|████▉ | 497/1000 [01:02<00:33, 15.21it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 50%|████▉ | 499/1000 [01:03<00:31, 15.85it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
sample: 50%|█████ | 502/1000 [01:03<00:29, 16.97it/s, 127 steps of size 7.37e-02. acc. prob=0.92]
sample: 50%|█████ | 504/1000 [01:03<00:30, 16.41it/s, 63 steps of size 7.37e-02. acc. prob=0.84]
sample: 51%|█████ | 507/1000 [01:03<00:27, 18.17it/s, 63 steps of size 7.37e-02. acc. prob=0.85]
sample: 51%|█████ | 510/1000 [01:03<00:25, 19.48it/s, 63 steps of size 7.37e-02. acc. prob=0.89]
sample: 51%|█████▏ | 513/1000 [01:03<00:22, 21.55it/s, 31 steps of size 7.37e-02. acc. prob=0.88]
sample: 52%|█████▏ | 515/1000 [01:03<00:25, 18.72it/s, 127 steps of size 7.37e-02. acc. prob=0.89]
sample: 52%|█████▏ | 518/1000 [01:03<00:24, 19.86it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 52%|█████▏ | 521/1000 [01:04<00:23, 20.67it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 52%|█████▏ | 524/1000 [01:04<00:22, 21.23it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 53%|█████▎ | 526/1000 [01:04<00:24, 19.42it/s, 127 steps of size 7.37e-02. acc. prob=0.90]
sample: 53%|█████▎ | 529/1000 [01:04<00:21, 21.55it/s, 31 steps of size 7.37e-02. acc. prob=0.89]
sample: 53%|█████▎ | 532/1000 [01:04<00:21, 21.92it/s, 63 steps of size 7.37e-02. acc. prob=0.89]
sample: 54%|█████▎ | 535/1000 [01:04<00:21, 22.13it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 54%|█████▍ | 538/1000 [01:04<00:20, 22.28it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 54%|█████▍ | 541/1000 [01:05<00:20, 22.34it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 54%|█████▍ | 544/1000 [01:05<00:20, 22.39it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 55%|█████▍ | 546/1000 [01:05<00:22, 20.25it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 55%|█████▍ | 548/1000 [01:05<00:24, 18.59it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 55%|█████▌ | 551/1000 [01:05<00:22, 19.80it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 55%|█████▌ | 554/1000 [01:05<00:21, 20.60it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 56%|█████▌ | 557/1000 [01:05<00:22, 19.28it/s, 127 steps of size 7.37e-02. acc. prob=0.90]
sample: 56%|█████▌ | 560/1000 [01:05<00:21, 20.13it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 56%|█████▌ | 562/1000 [01:06<00:23, 18.59it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 56%|█████▋ | 564/1000 [01:06<00:24, 17.48it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 57%|█████▋ | 566/1000 [01:06<00:25, 16.84it/s, 127 steps of size 7.37e-02. acc. prob=0.90]
sample: 57%|█████▋ | 569/1000 [01:06<00:24, 17.63it/s, 95 steps of size 7.37e-02. acc. prob=0.91]
sample: 57%|█████▋ | 572/1000 [01:06<00:22, 19.10it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 57%|█████▊ | 575/1000 [01:06<00:21, 20.09it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 58%|█████▊ | 578/1000 [01:06<00:20, 20.89it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 58%|█████▊ | 581/1000 [01:07<00:18, 22.53it/s, 31 steps of size 7.37e-02. acc. prob=0.91]
sample: 58%|█████▊ | 584/1000 [01:07<00:18, 22.55it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 59%|█████▊ | 587/1000 [01:07<00:18, 22.68it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 59%|█████▉ | 589/1000 [01:07<00:20, 20.45it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 59%|█████▉ | 592/1000 [01:07<00:19, 21.12it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 59%|█████▉ | 594/1000 [01:07<00:21, 19.31it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 60%|█████▉ | 597/1000 [01:07<00:19, 20.31it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 60%|██████ | 600/1000 [01:07<00:19, 20.94it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 60%|██████ | 603/1000 [01:08<00:18, 21.47it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 61%|██████ | 606/1000 [01:08<00:19, 19.81it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 61%|██████ | 608/1000 [01:08<00:21, 18.48it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 61%|██████ | 611/1000 [01:08<00:19, 19.65it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 61%|██████▏ | 614/1000 [01:08<00:18, 20.51it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 62%|██████▏ | 617/1000 [01:08<00:19, 19.20it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 62%|██████▏ | 619/1000 [01:09<00:21, 18.03it/s, 127 steps of size 7.37e-02. acc. prob=0.90]
sample: 62%|██████▏ | 621/1000 [01:09<00:21, 18.03it/s, 31 steps of size 7.37e-02. acc. prob=0.90]
sample: 62%|██████▏ | 623/1000 [01:09<00:21, 17.17it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 63%|██████▎ | 626/1000 [01:09<00:20, 18.70it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 63%|██████▎ | 628/1000 [01:09<00:20, 18.51it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 63%|██████▎ | 631/1000 [01:09<00:18, 19.75it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 63%|██████▎ | 634/1000 [01:09<00:17, 20.56it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 64%|██████▎ | 637/1000 [01:09<00:16, 22.36it/s, 31 steps of size 7.37e-02. acc. prob=0.90]
sample: 64%|██████▍ | 640/1000 [01:10<00:15, 22.51it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 64%|██████▍ | 643/1000 [01:10<00:14, 23.89it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 65%|██████▍ | 646/1000 [01:10<00:15, 23.52it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 65%|██████▍ | 648/1000 [01:10<00:16, 20.89it/s, 127 steps of size 7.37e-02. acc. prob=0.90]
sample: 65%|██████▌ | 651/1000 [01:10<00:15, 22.61it/s, 31 steps of size 7.37e-02. acc. prob=0.90]
sample: 65%|██████▌ | 653/1000 [01:10<00:16, 21.41it/s, 95 steps of size 7.37e-02. acc. prob=0.90]
sample: 66%|██████▌ | 656/1000 [01:10<00:15, 21.84it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 66%|██████▌ | 659/1000 [01:10<00:15, 22.11it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 66%|██████▌ | 662/1000 [01:10<00:15, 22.23it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 66%|██████▋ | 665/1000 [01:11<00:14, 22.38it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 67%|██████▋ | 667/1000 [01:11<00:16, 20.05it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 67%|██████▋ | 670/1000 [01:11<00:15, 20.80it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 67%|██████▋ | 673/1000 [01:11<00:15, 21.31it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 68%|██████▊ | 676/1000 [01:11<00:14, 22.88it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 68%|██████▊ | 679/1000 [01:11<00:13, 24.09it/s, 31 steps of size 7.37e-02. acc. prob=0.90]
sample: 68%|██████▊ | 681/1000 [01:11<00:14, 21.34it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 68%|██████▊ | 684/1000 [01:12<00:14, 21.67it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 69%|██████▊ | 687/1000 [01:12<00:13, 23.08it/s, 31 steps of size 7.37e-02. acc. prob=0.90]
sample: 69%|██████▉ | 689/1000 [01:12<00:14, 21.69it/s, 31 steps of size 7.37e-02. acc. prob=0.90]
sample: 69%|██████▉ | 691/1000 [01:12<00:15, 19.38it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 69%|██████▉ | 694/1000 [01:12<00:15, 20.16it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 70%|██████▉ | 697/1000 [01:12<00:14, 20.69it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 70%|██████▉ | 699/1000 [01:12<00:15, 18.93it/s, 63 steps of size 7.37e-02. acc. prob=0.90]
sample: 70%|███████ | 702/1000 [01:12<00:16, 18.21it/s, 127 steps of size 7.37e-02. acc. prob=0.90]
sample: 70%|███████ | 705/1000 [01:13<00:15, 19.38it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 71%|███████ | 708/1000 [01:13<00:13, 21.25it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 71%|███████ | 711/1000 [01:13<00:13, 21.60it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 71%|███████▏ | 714/1000 [01:13<00:13, 21.85it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 72%|███████▏ | 717/1000 [01:13<00:12, 22.09it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 72%|███████▏ | 720/1000 [01:13<00:11, 23.34it/s, 31 steps of size 7.37e-02. acc. prob=0.91]
sample: 72%|███████▏ | 723/1000 [01:13<00:12, 23.07it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 73%|███████▎ | 726/1000 [01:13<00:11, 23.11it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 73%|███████▎ | 729/1000 [01:14<00:11, 22.95it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 73%|███████▎ | 732/1000 [01:14<00:11, 22.88it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 73%|███████▎ | 734/1000 [01:14<00:12, 20.50it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 74%|███████▎ | 736/1000 [01:14<00:15, 17.18it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 74%|███████▍ | 739/1000 [01:14<00:13, 18.74it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 74%|███████▍ | 741/1000 [01:14<00:14, 17.80it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 74%|███████▍ | 743/1000 [01:14<00:15, 16.94it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 74%|███████▍ | 745/1000 [01:15<00:14, 17.26it/s, 31 steps of size 7.37e-02. acc. prob=0.91]
sample: 75%|███████▍ | 747/1000 [01:15<00:15, 16.67it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 75%|███████▌ | 750/1000 [01:15<00:13, 18.50it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 75%|███████▌ | 753/1000 [01:15<00:12, 19.77it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 76%|███████▌ | 756/1000 [01:15<00:11, 20.65it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 76%|███████▌ | 758/1000 [01:15<00:12, 18.93it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 76%|███████▌ | 761/1000 [01:15<00:13, 18.22it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 76%|███████▋ | 763/1000 [01:16<00:12, 18.24it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 77%|███████▋ | 766/1000 [01:16<00:11, 19.56it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 77%|███████▋ | 769/1000 [01:16<00:12, 18.64it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 77%|███████▋ | 771/1000 [01:16<00:14, 16.11it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 77%|███████▋ | 773/1000 [01:16<00:14, 15.88it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 78%|███████▊ | 776/1000 [01:16<00:12, 17.80it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 78%|███████▊ | 778/1000 [01:16<00:13, 16.99it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 78%|███████▊ | 781/1000 [01:17<00:11, 18.69it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 78%|███████▊ | 783/1000 [01:17<00:11, 18.51it/s, 31 steps of size 7.37e-02. acc. prob=0.91]
sample: 79%|███████▊ | 786/1000 [01:17<00:10, 19.80it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 79%|███████▉ | 789/1000 [01:17<00:10, 20.66it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 79%|███████▉ | 792/1000 [01:17<00:09, 21.31it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 79%|███████▉ | 794/1000 [01:17<00:10, 20.50it/s, 95 steps of size 7.37e-02. acc. prob=0.91]
sample: 80%|███████▉ | 797/1000 [01:17<00:09, 21.25it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 80%|████████ | 800/1000 [01:17<00:09, 21.76it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 80%|████████ | 803/1000 [01:18<00:08, 22.05it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 81%|████████ | 806/1000 [01:18<00:08, 22.25it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 81%|████████ | 809/1000 [01:18<00:08, 22.39it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 81%|████████ | 811/1000 [01:18<00:09, 20.18it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 81%|████████▏ | 813/1000 [01:18<00:10, 18.67it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 82%|████████▏ | 815/1000 [01:18<00:10, 17.56it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 82%|████████▏ | 818/1000 [01:18<00:09, 19.12it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 82%|████████▏ | 821/1000 [01:18<00:08, 20.26it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 82%|████████▏ | 823/1000 [01:19<00:09, 18.67it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 82%|████████▎ | 825/1000 [01:19<00:09, 17.57it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 83%|████████▎ | 827/1000 [01:19<00:10, 16.92it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 83%|████████▎ | 829/1000 [01:19<00:10, 16.38it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 83%|████████▎ | 832/1000 [01:19<00:10, 16.74it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 84%|████████▎ | 835/1000 [01:19<00:09, 18.28it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 84%|████████▍ | 838/1000 [01:19<00:08, 19.54it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 84%|████████▍ | 840/1000 [01:20<00:08, 18.25it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 84%|████████▍ | 843/1000 [01:20<00:08, 17.78it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 85%|████████▍ | 846/1000 [01:20<00:08, 19.12it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 85%|████████▍ | 849/1000 [01:20<00:07, 20.13it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 85%|████████▌ | 852/1000 [01:20<00:07, 20.88it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 86%|████████▌ | 855/1000 [01:20<00:06, 21.44it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 86%|████████▌ | 858/1000 [01:20<00:06, 21.79it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 86%|████████▌ | 860/1000 [01:21<00:07, 19.80it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 86%|████████▋ | 863/1000 [01:21<00:06, 21.70it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 87%|████████▋ | 866/1000 [01:21<00:06, 21.94it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 87%|████████▋ | 869/1000 [01:21<00:05, 22.07it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 87%|████████▋ | 872/1000 [01:21<00:05, 22.13it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 88%|████████▊ | 875/1000 [01:21<00:05, 22.30it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 88%|████████▊ | 877/1000 [01:21<00:06, 20.06it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 88%|████████▊ | 880/1000 [01:21<00:05, 21.82it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 88%|████████▊ | 883/1000 [01:22<00:05, 22.01it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 89%|████████▊ | 886/1000 [01:22<00:04, 23.33it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 89%|████████▉ | 888/1000 [01:22<00:05, 20.61it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 89%|████████▉ | 891/1000 [01:22<00:05, 20.97it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 89%|████████▉ | 893/1000 [01:22<00:05, 19.89it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 90%|████████▉ | 895/1000 [01:22<00:06, 16.47it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 90%|████████▉ | 898/1000 [01:22<00:05, 18.04it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 90%|█████████ | 900/1000 [01:23<00:05, 17.09it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 90%|█████████ | 903/1000 [01:23<00:05, 18.61it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 91%|█████████ | 907/1000 [01:23<00:04, 21.86it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 91%|█████████ | 910/1000 [01:23<00:04, 21.98it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 91%|█████████▏| 913/1000 [01:23<00:03, 22.14it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 92%|█████████▏| 916/1000 [01:23<00:03, 22.26it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 92%|█████████▏| 919/1000 [01:23<00:03, 22.29it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 92%|█████████▏| 922/1000 [01:23<00:03, 22.33it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 92%|█████████▎| 925/1000 [01:24<00:03, 22.33it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 93%|█████████▎| 928/1000 [01:24<00:03, 23.56it/s, 31 steps of size 7.37e-02. acc. prob=0.91]
sample: 93%|█████████▎| 931/1000 [01:24<00:02, 23.27it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 93%|█████████▎| 934/1000 [01:24<00:02, 23.21it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 94%|█████████▎| 937/1000 [01:24<00:02, 23.05it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 94%|█████████▍| 940/1000 [01:24<00:02, 23.05it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 94%|█████████▍| 943/1000 [01:24<00:02, 22.99it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 95%|█████████▍| 946/1000 [01:24<00:02, 24.15it/s, 31 steps of size 7.37e-02. acc. prob=0.91]
sample: 95%|█████████▍| 948/1000 [01:25<00:02, 21.37it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 95%|█████████▌| 951/1000 [01:25<00:02, 22.97it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 95%|█████████▌| 954/1000 [01:25<00:02, 20.66it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 96%|█████████▌| 957/1000 [01:25<00:02, 21.20it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 96%|█████████▌| 960/1000 [01:25<00:01, 21.63it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 96%|█████████▌| 962/1000 [01:25<00:01, 19.68it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 96%|█████████▋| 965/1000 [01:25<00:01, 20.56it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 97%|█████████▋| 967/1000 [01:26<00:01, 18.89it/s, 127 steps of size 7.37e-02. acc. prob=0.91]
sample: 97%|█████████▋| 970/1000 [01:26<00:01, 19.99it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 97%|█████████▋| 973/1000 [01:26<00:01, 20.77it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 98%|█████████▊| 976/1000 [01:26<00:01, 21.34it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 98%|█████████▊| 979/1000 [01:26<00:00, 21.72it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 98%|█████████▊| 982/1000 [01:26<00:00, 21.99it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 98%|█████████▊| 985/1000 [01:26<00:00, 23.48it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 99%|█████████▊| 987/1000 [01:26<00:00, 20.79it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 99%|█████████▉| 990/1000 [01:27<00:00, 22.46it/s, 31 steps of size 7.37e-02. acc. prob=0.91]
sample: 99%|█████████▉| 993/1000 [01:27<00:00, 23.89it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 100%|█████████▉| 996/1000 [01:27<00:00, 23.39it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 100%|█████████▉| 998/1000 [01:27<00:00, 20.75it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
sample: 100%|██████████| 1000/1000 [01:27<00:00, 11.43it/s, 63 steps of size 7.37e-02. acc. prob=0.91]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 0%| | 0/1000 [00:13<?, ?it/s, 1023 steps of size 2.44e-03. acc. prob=0.73]
warmup: 5%|▌ | 50/1000 [00:13<04:22, 3.62it/s, 1023 steps of size 4.13e-03. acc. prob=0.73]
warmup: 5%|▌ | 51/1000 [00:14<04:33, 3.47it/s, 1023 steps of size 4.47e-03. acc. prob=0.74]
warmup: 5%|▌ | 52/1000 [00:14<04:35, 3.44it/s, 511 steps of size 7.11e-03. acc. prob=0.74]
warmup: 5%|▌ | 53/1000 [00:15<04:38, 3.40it/s, 511 steps of size 5.38e-03. acc. prob=0.74]
warmup: 5%|▌ | 54/1000 [00:15<04:42, 3.35it/s, 511 steps of size 8.62e-03. acc. prob=0.74]
warmup: 6%|▌ | 55/1000 [00:15<04:32, 3.47it/s, 255 steps of size 9.17e-03. acc. prob=0.74]
warmup: 6%|▌ | 57/1000 [00:16<04:45, 3.31it/s, 1023 steps of size 2.51e-03. acc. prob=0.74]
warmup: 6%|▌ | 58/1000 [00:17<05:32, 2.83it/s, 1023 steps of size 4.05e-03. acc. prob=0.74]
warmup: 6%|▌ | 59/1000 [00:17<06:23, 2.45it/s, 1023 steps of size 5.43e-03. acc. prob=0.74]
warmup: 6%|▌ | 60/1000 [00:18<06:13, 2.52it/s, 511 steps of size 7.42e-03. acc. prob=0.75]
warmup: 6%|▌ | 61/1000 [00:18<06:04, 2.58it/s, 511 steps of size 1.95e-03. acc. prob=0.74]
warmup: 6%|▌ | 62/1000 [00:19<07:11, 2.17it/s, 1023 steps of size 3.17e-03. acc. prob=0.74]
warmup: 6%|▋ | 63/1000 [00:19<08:06, 1.93it/s, 1023 steps of size 3.95e-03. acc. prob=0.74]
warmup: 6%|▋ | 64/1000 [00:20<08:51, 1.76it/s, 1023 steps of size 5.28e-03. acc. prob=0.75]
warmup: 6%|▋ | 65/1000 [00:21<07:59, 1.95it/s, 511 steps of size 7.05e-03. acc. prob=0.75]
warmup: 7%|▋ | 66/1000 [00:21<07:21, 2.11it/s, 511 steps of size 3.80e-03. acc. prob=0.74]
warmup: 7%|▋ | 67/1000 [00:22<08:30, 1.83it/s, 1023 steps of size 5.80e-03. acc. prob=0.75]
warmup: 7%|▋ | 68/1000 [00:22<07:40, 2.02it/s, 511 steps of size 5.73e-03. acc. prob=0.75]
warmup: 7%|▋ | 69/1000 [00:22<07:06, 2.18it/s, 511 steps of size 8.99e-03. acc. prob=0.75]
warmup: 7%|▋ | 71/1000 [00:23<06:32, 2.37it/s, 1023 steps of size 2.79e-03. acc. prob=0.74]
warmup: 7%|▋ | 72/1000 [00:24<07:45, 1.99it/s, 1023 steps of size 4.39e-03. acc. prob=0.75]
warmup: 7%|▋ | 73/1000 [00:25<08:42, 1.77it/s, 1023 steps of size 6.74e-03. acc. prob=0.75]
warmup: 7%|▋ | 74/1000 [00:25<07:51, 1.96it/s, 511 steps of size 5.28e-03. acc. prob=0.75]
warmup: 8%|▊ | 75/1000 [00:25<07:13, 2.14it/s, 511 steps of size 7.86e-03. acc. prob=0.75]
warmup: 8%|▊ | 76/1000 [00:26<06:42, 2.30it/s, 511 steps of size 3.61e-03. acc. prob=0.75]
warmup: 8%|▊ | 77/1000 [00:26<07:53, 1.95it/s, 1023 steps of size 4.48e-03. acc. prob=0.75]
warmup: 8%|▊ | 78/1000 [00:27<08:44, 1.76it/s, 1023 steps of size 3.71e-03. acc. prob=0.75]
warmup: 8%|▊ | 79/1000 [00:28<09:20, 1.64it/s, 1023 steps of size 5.40e-03. acc. prob=0.75]
warmup: 8%|▊ | 80/1000 [00:28<08:08, 1.88it/s, 511 steps of size 8.30e-03. acc. prob=0.76]
warmup: 8%|▊ | 82/1000 [00:29<07:10, 2.13it/s, 1023 steps of size 4.11e-03. acc. prob=0.75]
warmup: 8%|▊ | 83/1000 [00:30<08:03, 1.90it/s, 1023 steps of size 5.23e-03. acc. prob=0.75]
warmup: 8%|▊ | 84/1000 [00:30<08:48, 1.73it/s, 1023 steps of size 7.47e-03. acc. prob=0.76]
warmup: 8%|▊ | 85/1000 [00:31<07:57, 1.92it/s, 511 steps of size 7.72e-03. acc. prob=0.76]
warmup: 9%|▊ | 86/1000 [00:31<06:46, 2.25it/s, 347 steps of size 3.76e-03. acc. prob=0.75]
warmup: 9%|▊ | 87/1000 [00:32<07:48, 1.95it/s, 1023 steps of size 5.70e-03. acc. prob=0.76]
warmup: 9%|▉ | 88/1000 [00:32<07:06, 2.14it/s, 511 steps of size 5.70e-03. acc. prob=0.76]
warmup: 9%|▉ | 89/1000 [00:32<05:47, 2.62it/s, 255 steps of size 8.39e-03. acc. prob=0.76]
warmup: 9%|▉ | 91/1000 [00:33<05:55, 2.56it/s, 1023 steps of size 3.43e-03. acc. prob=0.75]
warmup: 9%|▉ | 92/1000 [00:34<07:02, 2.15it/s, 1023 steps of size 5.17e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:34<06:35, 2.30it/s, 511 steps of size 6.21e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:34<06:13, 2.43it/s, 511 steps of size 7.80e-03. acc. prob=0.76]
warmup: 10%|▉ | 95/1000 [00:35<05:58, 2.53it/s, 511 steps of size 1.11e-02. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:35<05:39, 2.66it/s, 1023 steps of size 3.71e-03. acc. prob=0.76]
warmup: 10%|▉ | 98/1000 [00:36<06:51, 2.19it/s, 1023 steps of size 5.05e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:37<06:27, 2.32it/s, 511 steps of size 7.49e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:37<05:26, 2.76it/s, 255 steps of size 7.23e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:38<07:09, 2.09it/s, 1023 steps of size 1.04e-01. acc. prob=0.76]
warmup: 10%|█ | 105/1000 [00:38<02:59, 4.98it/s, 3 steps of size 2.32e-02. acc. prob=0.76]
warmup: 11%|█ | 106/1000 [00:38<02:43, 5.47it/s, 127 steps of size 4.07e-02. acc. prob=0.76]
warmup: 11%|█ | 107/1000 [00:38<02:28, 6.02it/s, 127 steps of size 7.05e-02. acc. prob=0.76]
warmup: 11%|█ | 109/1000 [00:38<01:50, 8.08it/s, 63 steps of size 2.93e-02. acc. prob=0.76]
warmup: 11%|█ | 110/1000 [00:38<02:04, 7.16it/s, 255 steps of size 5.47e-02. acc. prob=0.76]
warmup: 11%|█ | 111/1000 [00:38<01:56, 7.66it/s, 127 steps of size 1.01e-01. acc. prob=0.77]
warmup: 11%|█▏ | 114/1000 [00:38<01:33, 9.51it/s, 255 steps of size 2.88e-02. acc. prob=0.76]
warmup: 12%|█▏ | 116/1000 [00:39<01:24, 10.45it/s, 63 steps of size 8.75e-02. acc. prob=0.77]
warmup: 12%|█▏ | 119/1000 [00:39<01:07, 13.00it/s, 63 steps of size 1.26e-01. acc. prob=0.77]
warmup: 12%|█▏ | 121/1000 [00:39<01:07, 13.11it/s, 127 steps of size 8.64e-02. acc. prob=0.77]
warmup: 12%|█▏ | 124/1000 [00:39<00:55, 15.81it/s, 63 steps of size 8.67e-02. acc. prob=0.77]
warmup: 13%|█▎ | 126/1000 [00:39<00:55, 15.85it/s, 95 steps of size 1.74e-01. acc. prob=0.77]
warmup: 13%|█▎ | 129/1000 [00:39<00:48, 18.13it/s, 63 steps of size 1.63e-01. acc. prob=0.77]
warmup: 13%|█▎ | 132/1000 [00:39<00:43, 19.83it/s, 63 steps of size 2.23e-01. acc. prob=0.77]
warmup: 13%|█▎ | 134/1000 [00:40<00:54, 15.96it/s, 255 steps of size 4.01e-02. acc. prob=0.77]
warmup: 14%|█▎ | 136/1000 [00:40<00:56, 15.27it/s, 63 steps of size 1.08e-01. acc. prob=0.77]
warmup: 14%|█▍ | 140/1000 [00:40<00:57, 15.04it/s, 255 steps of size 4.41e-02. acc. prob=0.77]
warmup: 14%|█▍ | 141/1000 [00:40<01:01, 14.05it/s, 127 steps of size 7.04e-02. acc. prob=0.77]
warmup: 14%|█▍ | 144/1000 [00:40<00:54, 15.75it/s, 63 steps of size 5.60e-02. acc. prob=0.77]
warmup: 15%|█▍ | 147/1000 [00:40<00:49, 17.09it/s, 63 steps of size 1.28e-01. acc. prob=0.77]
warmup: 15%|█▌ | 150/1000 [00:41<00:47, 17.89it/s, 127 steps of size 5.81e-02. acc. prob=0.77]
warmup: 15%|█▌ | 153/1000 [00:41<00:43, 19.53it/s, 95 steps of size 1.60e-02. acc. prob=0.76]
warmup: 15%|█▌ | 154/1000 [00:41<00:59, 14.24it/s, 255 steps of size 2.19e-02. acc. prob=0.77]
warmup: 16%|█▌ | 155/1000 [00:41<01:15, 11.14it/s, 255 steps of size 3.41e-02. acc. prob=0.77]
warmup: 16%|█▌ | 156/1000 [00:41<01:31, 9.18it/s, 255 steps of size 5.67e-02. acc. prob=0.77]
warmup: 16%|█▌ | 158/1000 [00:41<01:21, 10.34it/s, 63 steps of size 1.54e-01. acc. prob=0.77]
warmup: 16%|█▌ | 160/1000 [00:42<01:48, 7.76it/s, 511 steps of size 2.27e-02. acc. prob=0.77]
warmup: 16%|█▌ | 161/1000 [00:42<01:58, 7.09it/s, 255 steps of size 4.15e-02. acc. prob=0.77]
warmup: 16%|█▌ | 162/1000 [00:42<02:07, 6.57it/s, 255 steps of size 7.62e-02. acc. prob=0.77]
warmup: 16%|█▋ | 165/1000 [00:42<01:19, 10.45it/s, 5 steps of size 2.35e-02. acc. prob=0.77]
warmup: 17%|█▋ | 166/1000 [00:43<01:35, 8.69it/s, 255 steps of size 3.86e-02. acc. prob=0.77]
warmup: 17%|█▋ | 167/1000 [00:43<01:33, 8.92it/s, 127 steps of size 6.34e-02. acc. prob=0.77]
warmup: 17%|█▋ | 169/1000 [00:43<01:14, 11.21it/s, 63 steps of size 1.32e-01. acc. prob=0.77]
warmup: 17%|█▋ | 171/1000 [00:43<01:18, 10.62it/s, 255 steps of size 2.30e-02. acc. prob=0.77]
warmup: 17%|█▋ | 172/1000 [00:43<02:08, 6.44it/s, 511 steps of size 3.96e-02. acc. prob=0.77]
warmup: 17%|█▋ | 174/1000 [00:43<01:43, 7.99it/s, 63 steps of size 9.73e-02. acc. prob=0.77]
warmup: 18%|█▊ | 177/1000 [00:44<01:26, 9.46it/s, 255 steps of size 3.05e-02. acc. prob=0.77]
warmup: 18%|█▊ | 179/1000 [00:44<01:18, 10.41it/s, 63 steps of size 9.35e-02. acc. prob=0.77]
warmup: 18%|█▊ | 181/1000 [00:44<01:15, 10.80it/s, 127 steps of size 4.44e-02. acc. prob=0.77]
warmup: 18%|█▊ | 183/1000 [00:44<01:16, 10.69it/s, 127 steps of size 1.15e-01. acc. prob=0.77]
warmup: 18%|█▊ | 185/1000 [00:44<01:18, 10.32it/s, 255 steps of size 2.82e-02. acc. prob=0.77]
warmup: 19%|█▊ | 186/1000 [00:45<01:33, 8.75it/s, 255 steps of size 4.86e-02. acc. prob=0.77]
warmup: 19%|█▉ | 188/1000 [00:45<01:21, 9.99it/s, 63 steps of size 5.89e-02. acc. prob=0.77]
warmup: 19%|█▉ | 191/1000 [00:45<01:03, 12.77it/s, 63 steps of size 1.22e-01. acc. prob=0.78]
warmup: 19%|█▉ | 193/1000 [00:45<01:07, 11.91it/s, 255 steps of size 2.82e-02. acc. prob=0.77]
warmup: 20%|█▉ | 195/1000 [00:45<01:10, 11.44it/s, 127 steps of size 7.15e-02. acc. prob=0.77]
warmup: 20%|█▉ | 198/1000 [00:45<00:57, 13.90it/s, 63 steps of size 9.55e-02. acc. prob=0.78]
warmup: 20%|██ | 201/1000 [00:46<00:53, 14.99it/s, 127 steps of size 9.20e-02. acc. prob=0.78]
warmup: 20%|██ | 204/1000 [00:46<00:48, 16.43it/s, 127 steps of size 5.95e-02. acc. prob=0.78]
warmup: 21%|██ | 206/1000 [00:46<00:50, 15.69it/s, 63 steps of size 1.08e-01. acc. prob=0.78]
warmup: 21%|██ | 209/1000 [00:46<00:53, 14.91it/s, 255 steps of size 4.40e-02. acc. prob=0.77]
warmup: 21%|██ | 211/1000 [00:46<00:58, 13.50it/s, 127 steps of size 7.03e-02. acc. prob=0.78]
warmup: 21%|██▏ | 213/1000 [00:46<00:57, 13.59it/s, 127 steps of size 9.44e-02. acc. prob=0.78]
warmup: 22%|██▏ | 216/1000 [00:47<00:48, 16.25it/s, 63 steps of size 5.76e-02. acc. prob=0.78]
warmup: 22%|██▏ | 218/1000 [00:47<00:47, 16.41it/s, 31 steps of size 9.52e-02. acc. prob=0.78]
warmup: 22%|██▏ | 221/1000 [00:47<00:46, 16.88it/s, 127 steps of size 5.34e-02. acc. prob=0.78]
warmup: 22%|██▏ | 223/1000 [00:47<00:48, 16.02it/s, 63 steps of size 3.67e-02. acc. prob=0.78]
warmup: 22%|██▏ | 224/1000 [00:47<00:57, 13.38it/s, 191 steps of size 5.58e-02. acc. prob=0.78]
warmup: 23%|██▎ | 226/1000 [00:47<00:54, 14.24it/s, 31 steps of size 1.23e-01. acc. prob=0.78]
warmup: 23%|██▎ | 228/1000 [00:47<00:49, 15.47it/s, 127 steps of size 5.73e-02. acc. prob=0.78]
warmup: 23%|██▎ | 230/1000 [00:48<00:56, 13.55it/s, 127 steps of size 6.74e-02. acc. prob=0.78]
warmup: 23%|██▎ | 233/1000 [00:48<00:48, 15.71it/s, 63 steps of size 1.24e-01. acc. prob=0.78]
warmup: 24%|██▎ | 236/1000 [00:48<00:40, 18.88it/s, 63 steps of size 9.57e-02. acc. prob=0.78]
warmup: 24%|██▍ | 238/1000 [00:48<00:41, 18.31it/s, 127 steps of size 9.44e-02. acc. prob=0.78]
warmup: 24%|██▍ | 241/1000 [00:48<00:39, 19.04it/s, 63 steps of size 3.69e-02. acc. prob=0.78]
warmup: 24%|██▍ | 243/1000 [00:48<00:43, 17.46it/s, 63 steps of size 6.35e-02. acc. prob=0.78]
warmup: 25%|██▍ | 247/1000 [00:48<00:47, 15.87it/s, 255 steps of size 5.32e-02. acc. prob=0.78]
warmup: 25%|██▍ | 249/1000 [00:49<00:48, 15.60it/s, 127 steps of size 4.06e-02. acc. prob=0.78]
warmup: 25%|██▌ | 251/1000 [00:49<00:49, 15.27it/s, 63 steps of size 7.77e-01. acc. prob=0.78]
warmup: 26%|██▌ | 255/1000 [00:49<00:39, 18.88it/s, 95 steps of size 1.01e-01. acc. prob=0.78]
warmup: 26%|██▌ | 257/1000 [00:49<00:42, 17.55it/s, 127 steps of size 8.93e-02. acc. prob=0.78]
warmup: 26%|██▌ | 260/1000 [00:49<00:47, 15.68it/s, 255 steps of size 2.30e-02. acc. prob=0.78]
warmup: 26%|██▌ | 261/1000 [00:49<00:59, 12.44it/s, 255 steps of size 4.28e-02. acc. prob=0.78]
warmup: 26%|██▋ | 263/1000 [00:50<00:56, 12.97it/s, 63 steps of size 1.45e-01. acc. prob=0.78]
warmup: 27%|██▋ | 267/1000 [00:50<00:40, 17.98it/s, 63 steps of size 1.33e-01. acc. prob=0.78]
warmup: 27%|██▋ | 269/1000 [00:50<00:47, 15.47it/s, 127 steps of size 6.03e-02. acc. prob=0.78]
warmup: 27%|██▋ | 272/1000 [00:50<00:40, 18.02it/s, 31 steps of size 2.39e-02. acc. prob=0.78]
warmup: 27%|██▋ | 273/1000 [00:50<00:53, 13.58it/s, 255 steps of size 4.34e-02. acc. prob=0.78]
warmup: 28%|██▊ | 275/1000 [00:50<00:52, 13.72it/s, 63 steps of size 9.31e-02. acc. prob=0.78]
warmup: 28%|██▊ | 277/1000 [00:50<00:52, 13.90it/s, 127 steps of size 7.98e-02. acc. prob=0.78]
warmup: 28%|██▊ | 280/1000 [00:51<00:42, 16.96it/s, 63 steps of size 9.77e-02. acc. prob=0.78]
warmup: 28%|██▊ | 283/1000 [00:51<00:51, 14.00it/s, 255 steps of size 5.94e-02. acc. prob=0.78]
warmup: 28%|██▊ | 285/1000 [00:51<00:55, 12.89it/s, 127 steps of size 1.11e-01. acc. prob=0.78]
warmup: 29%|██▊ | 287/1000 [00:51<00:51, 13.74it/s, 127 steps of size 7.75e-02. acc. prob=0.78]
warmup: 29%|██▉ | 291/1000 [00:51<00:43, 16.39it/s, 127 steps of size 6.36e-02. acc. prob=0.78]
warmup: 29%|██▉ | 294/1000 [00:51<00:38, 18.57it/s, 31 steps of size 1.70e-01. acc. prob=0.78]
warmup: 30%|██▉ | 297/1000 [00:52<00:36, 19.21it/s, 63 steps of size 6.86e-02. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:52<00:33, 20.93it/s, 31 steps of size 1.56e-01. acc. prob=0.78]
warmup: 30%|███ | 302/1000 [00:52<00:35, 19.91it/s, 127 steps of size 4.51e-02. acc. prob=0.78]
warmup: 30%|███ | 304/1000 [00:52<00:38, 18.18it/s, 63 steps of size 1.03e-01. acc. prob=0.78]
warmup: 31%|███ | 307/1000 [00:52<00:36, 19.11it/s, 63 steps of size 1.30e-01. acc. prob=0.78]
warmup: 31%|███ | 311/1000 [00:52<00:37, 18.25it/s, 255 steps of size 4.57e-02. acc. prob=0.78]
warmup: 31%|███▏ | 313/1000 [00:53<00:39, 17.19it/s, 63 steps of size 9.58e-02. acc. prob=0.78]
warmup: 32%|███▏ | 316/1000 [00:53<00:37, 18.25it/s, 127 steps of size 6.05e-02. acc. prob=0.78]
warmup: 32%|███▏ | 319/1000 [00:53<00:33, 20.16it/s, 31 steps of size 1.32e-01. acc. prob=0.78]
warmup: 32%|███▏ | 322/1000 [00:53<00:39, 17.10it/s, 191 steps of size 6.10e-02. acc. prob=0.78]
warmup: 32%|███▎ | 325/1000 [00:53<00:36, 18.24it/s, 63 steps of size 1.57e-01. acc. prob=0.78]
warmup: 33%|███▎ | 327/1000 [00:53<00:37, 17.90it/s, 127 steps of size 4.58e-02. acc. prob=0.78]
warmup: 33%|███▎ | 330/1000 [00:53<00:33, 19.90it/s, 31 steps of size 1.22e-01. acc. prob=0.78]
warmup: 33%|███▎ | 333/1000 [00:54<00:34, 19.48it/s, 127 steps of size 7.23e-02. acc. prob=0.78]
warmup: 34%|███▎ | 336/1000 [00:54<00:33, 20.05it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 34%|███▍ | 339/1000 [00:54<00:32, 20.46it/s, 127 steps of size 5.48e-02. acc. prob=0.78]
warmup: 34%|███▍ | 341/1000 [00:54<00:35, 18.65it/s, 63 steps of size 9.50e-02. acc. prob=0.78]
warmup: 34%|███▍ | 343/1000 [00:54<00:37, 17.40it/s, 63 steps of size 7.63e-02. acc. prob=0.78]
warmup: 35%|███▍ | 347/1000 [00:54<00:34, 18.91it/s, 127 steps of size 5.67e-02. acc. prob=0.78]
warmup: 35%|███▍ | 349/1000 [00:54<00:36, 17.65it/s, 63 steps of size 9.45e-02. acc. prob=0.78]
warmup: 35%|███▌ | 352/1000 [00:55<00:34, 18.71it/s, 63 steps of size 8.01e-02. acc. prob=0.78]
warmup: 36%|███▌ | 355/1000 [00:55<00:32, 19.58it/s, 63 steps of size 1.34e-01. acc. prob=0.78]
warmup: 36%|███▌ | 359/1000 [00:55<00:28, 22.34it/s, 63 steps of size 7.44e-02. acc. prob=0.78]
warmup: 36%|███▌ | 362/1000 [00:55<00:27, 23.27it/s, 63 steps of size 1.41e-01. acc. prob=0.78]
warmup: 36%|███▋ | 365/1000 [00:55<00:27, 23.37it/s, 127 steps of size 7.04e-02. acc. prob=0.78]
warmup: 37%|███▋ | 368/1000 [00:55<00:27, 22.82it/s, 63 steps of size 9.27e-02. acc. prob=0.78]
warmup: 37%|███▋ | 372/1000 [00:55<00:25, 24.30it/s, 127 steps of size 6.92e-02. acc. prob=0.78]
warmup: 37%|███▋ | 374/1000 [00:56<00:29, 21.28it/s, 127 steps of size 8.91e-02. acc. prob=0.78]
warmup: 38%|███▊ | 377/1000 [00:56<00:27, 22.44it/s, 31 steps of size 9.17e-02. acc. prob=0.78]
warmup: 38%|███▊ | 381/1000 [00:56<00:25, 24.41it/s, 63 steps of size 9.66e-02. acc. prob=0.78]
warmup: 38%|███▊ | 385/1000 [00:56<00:26, 23.51it/s, 127 steps of size 7.93e-02. acc. prob=0.78]
warmup: 39%|███▉ | 388/1000 [00:56<00:25, 23.94it/s, 31 steps of size 8.88e-02. acc. prob=0.78]
warmup: 39%|███▉ | 391/1000 [00:56<00:24, 24.49it/s, 31 steps of size 1.83e-01. acc. prob=0.78]
warmup: 40%|███▉ | 395/1000 [00:56<00:23, 25.62it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 40%|███▉ | 399/1000 [00:57<00:24, 24.30it/s, 127 steps of size 5.75e-02. acc. prob=0.78]
warmup: 40%|████ | 400/1000 [00:57<00:30, 19.42it/s, 191 steps of size 7.86e-02. acc. prob=0.78]
warmup: 40%|████ | 403/1000 [00:57<00:28, 21.04it/s, 63 steps of size 1.29e-01. acc. prob=0.78]
warmup: 41%|████ | 406/1000 [00:57<00:26, 22.38it/s, 63 steps of size 7.72e-02. acc. prob=0.78]
warmup: 41%|████ | 409/1000 [00:57<00:25, 23.38it/s, 31 steps of size 1.29e-01. acc. prob=0.78]
warmup: 41%|████▏ | 413/1000 [00:57<00:23, 25.18it/s, 63 steps of size 1.31e-01. acc. prob=0.78]
warmup: 42%|████▏ | 417/1000 [00:57<00:21, 27.65it/s, 63 steps of size 9.73e-02. acc. prob=0.78]
warmup: 42%|████▏ | 420/1000 [00:57<00:21, 27.23it/s, 31 steps of size 7.46e-02. acc. prob=0.78]
warmup: 42%|████▏ | 422/1000 [00:58<00:25, 23.09it/s, 127 steps of size 9.84e-02. acc. prob=0.78]
warmup: 42%|████▎ | 425/1000 [00:58<00:23, 23.98it/s, 31 steps of size 1.54e-01. acc. prob=0.79]
warmup: 43%|████▎ | 427/1000 [00:58<00:25, 22.13it/s, 127 steps of size 6.28e-02. acc. prob=0.78]
warmup: 43%|████▎ | 430/1000 [00:58<00:25, 22.16it/s, 63 steps of size 8.65e-02. acc. prob=0.78]
warmup: 43%|████▎ | 433/1000 [00:58<00:24, 23.40it/s, 31 steps of size 1.70e-01. acc. prob=0.79]
warmup: 44%|████▎ | 436/1000 [00:58<00:27, 20.38it/s, 127 steps of size 6.62e-02. acc. prob=0.78]
warmup: 44%|████▍ | 440/1000 [00:58<00:24, 22.93it/s, 63 steps of size 5.48e-02. acc. prob=0.78]
warmup: 44%|████▍ | 441/1000 [00:58<00:33, 16.57it/s, 255 steps of size 6.66e-02. acc. prob=0.78]
warmup: 44%|████▍ | 444/1000 [00:59<00:30, 18.06it/s, 63 steps of size 9.10e-02. acc. prob=0.78]
warmup: 45%|████▍ | 447/1000 [00:59<00:27, 20.10it/s, 63 steps of size 6.22e-02. acc. prob=0.78]
warmup: 45%|████▍ | 449/1000 [00:59<00:29, 18.46it/s, 63 steps of size 5.08e-02. acc. prob=0.78]
warmup: 45%|████▌ | 451/1000 [00:59<00:31, 17.16it/s, 63 steps of size 9.28e-01. acc. prob=0.78]
warmup: 45%|████▌ | 454/1000 [00:59<00:33, 16.07it/s, 255 steps of size 2.08e-02. acc. prob=0.78]
warmup: 46%|████▌ | 455/1000 [00:59<00:43, 12.53it/s, 255 steps of size 3.25e-02. acc. prob=0.78]
warmup: 46%|████▌ | 456/1000 [01:00<00:53, 10.22it/s, 255 steps of size 5.48e-02. acc. prob=0.78]
warmup: 46%|████▌ | 458/1000 [01:00<00:51, 10.43it/s, 127 steps of size 1.32e-01. acc. prob=0.78]
warmup: 46%|████▌ | 460/1000 [01:00<00:53, 10.15it/s, 255 steps of size 3.70e-02. acc. prob=0.78]
warmup: 46%|████▌ | 462/1000 [01:00<00:48, 11.13it/s, 63 steps of size 9.11e-02. acc. prob=0.78]
warmup: 47%|████▋ | 466/1000 [01:00<00:34, 15.63it/s, 63 steps of size 1.48e-01. acc. prob=0.78]
warmup: 47%|████▋ | 468/1000 [01:00<00:39, 13.50it/s, 255 steps of size 3.19e-02. acc. prob=0.78]
warmup: 47%|████▋ | 470/1000 [01:01<00:36, 14.35it/s, 31 steps of size 9.20e-02. acc. prob=0.78]
warmup: 48%|████▊ | 475/1000 [01:01<00:33, 15.84it/s, 255 steps of size 3.40e-02. acc. prob=0.78]
warmup: 48%|████▊ | 477/1000 [01:01<00:33, 15.55it/s, 63 steps of size 9.81e-02. acc. prob=0.78]
warmup: 48%|████▊ | 481/1000 [01:01<00:33, 15.64it/s, 255 steps of size 3.87e-02. acc. prob=0.78]
warmup: 48%|████▊ | 483/1000 [01:01<00:33, 15.39it/s, 63 steps of size 9.29e-02. acc. prob=0.78]
warmup: 48%|████▊ | 485/1000 [01:02<00:32, 15.82it/s, 127 steps of size 6.35e-02. acc. prob=0.78]
warmup: 49%|████▉ | 488/1000 [01:02<00:29, 17.45it/s, 63 steps of size 2.25e-01. acc. prob=0.78]
warmup: 49%|████▉ | 491/1000 [01:02<00:30, 16.95it/s, 127 steps of size 8.29e-02. acc. prob=0.78]
warmup: 49%|████▉ | 494/1000 [01:02<00:26, 19.01it/s, 63 steps of size 1.39e-01. acc. prob=0.78]
warmup: 50%|████▉ | 497/1000 [01:02<00:25, 19.83it/s, 127 steps of size 4.22e-02. acc. prob=0.78]
warmup: 50%|████▉ | 499/1000 [01:02<00:27, 18.22it/s, 127 steps of size 9.41e-02. acc. prob=0.78]
sample: 50%|█████ | 502/1000 [01:02<00:25, 19.25it/s, 63 steps of size 7.44e-02. acc. prob=0.94]
sample: 50%|█████ | 505/1000 [01:03<00:24, 20.06it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 51%|█████ | 507/1000 [01:03<00:26, 18.45it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 51%|█████ | 510/1000 [01:03<00:25, 19.46it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 51%|█████▏ | 513/1000 [01:03<00:24, 20.14it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 52%|█████▏ | 516/1000 [01:03<00:25, 18.85it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 52%|█████▏ | 520/1000 [01:03<00:22, 21.78it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 52%|█████▏ | 523/1000 [01:03<00:21, 21.72it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 53%|█████▎ | 526/1000 [01:04<00:21, 21.67it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 53%|█████▎ | 528/1000 [01:04<00:24, 19.60it/s, 127 steps of size 7.44e-02. acc. prob=0.93]
sample: 53%|█████▎ | 531/1000 [01:04<00:23, 20.20it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 53%|█████▎ | 534/1000 [01:04<00:22, 20.71it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 54%|█████▎ | 536/1000 [01:04<00:24, 18.84it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 54%|█████▍ | 539/1000 [01:04<00:23, 19.68it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 54%|█████▍ | 541/1000 [01:04<00:25, 18.10it/s, 127 steps of size 7.44e-02. acc. prob=0.93]
sample: 54%|█████▍ | 544/1000 [01:04<00:23, 19.12it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 55%|█████▍ | 547/1000 [01:05<00:22, 19.96it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 55%|█████▌ | 550/1000 [01:05<00:21, 20.47it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 55%|█████▌ | 553/1000 [01:05<00:21, 20.94it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 56%|█████▌ | 556/1000 [01:05<00:19, 22.39it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 56%|█████▌ | 559/1000 [01:05<00:19, 22.28it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 56%|█████▌ | 562/1000 [01:05<00:19, 22.19it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 56%|█████▋ | 565/1000 [01:05<00:18, 23.24it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 57%|█████▋ | 568/1000 [01:06<00:18, 22.83it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 57%|█████▋ | 571/1000 [01:06<00:19, 22.53it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 57%|█████▋ | 574/1000 [01:06<00:18, 23.57it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 58%|█████▊ | 577/1000 [01:06<00:18, 22.98it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 58%|█████▊ | 580/1000 [01:06<00:17, 23.78it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 58%|█████▊ | 583/1000 [01:06<00:18, 23.11it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 59%|█████▊ | 586/1000 [01:06<00:18, 22.64it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 59%|█████▉ | 589/1000 [01:06<00:18, 22.31it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 59%|█████▉ | 592/1000 [01:07<00:18, 21.96it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 59%|█████▉ | 594/1000 [01:07<00:20, 19.61it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 60%|█████▉ | 597/1000 [01:07<00:20, 19.63it/s, 79 steps of size 7.44e-02. acc. prob=0.92]
sample: 60%|██████ | 600/1000 [01:07<00:19, 20.12it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 60%|██████ | 603/1000 [01:07<00:19, 20.53it/s, 63 steps of size 7.44e-02. acc. prob=0.91]
sample: 60%|██████ | 605/1000 [01:07<00:21, 18.72it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 61%|██████ | 608/1000 [01:07<00:19, 19.60it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 61%|██████ | 611/1000 [01:08<00:19, 20.11it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 61%|██████▏ | 614/1000 [01:08<00:18, 20.49it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 62%|██████▏ | 617/1000 [01:08<00:18, 20.69it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 62%|██████▏ | 620/1000 [01:08<00:18, 20.99it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 62%|██████▏ | 623/1000 [01:08<00:17, 21.12it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 62%|██████▎ | 625/1000 [01:08<00:19, 19.07it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 63%|██████▎ | 627/1000 [01:08<00:21, 17.71it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 63%|██████▎ | 630/1000 [01:09<00:19, 18.78it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 63%|██████▎ | 633/1000 [01:09<00:18, 19.76it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 64%|██████▎ | 636/1000 [01:09<00:17, 21.36it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 64%|██████▍ | 639/1000 [01:09<00:16, 21.59it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 64%|██████▍ | 642/1000 [01:09<00:16, 21.53it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 64%|██████▍ | 645/1000 [01:09<00:16, 21.69it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 65%|██████▍ | 647/1000 [01:09<00:18, 19.53it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 65%|██████▍ | 649/1000 [01:10<00:19, 18.00it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 65%|██████▌ | 652/1000 [01:10<00:17, 20.14it/s, 31 steps of size 7.44e-02. acc. prob=0.92]
sample: 65%|██████▌ | 654/1000 [01:10<00:17, 19.35it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 66%|██████▌ | 657/1000 [01:10<00:16, 21.29it/s, 31 steps of size 7.44e-02. acc. prob=0.92]
sample: 66%|██████▌ | 659/1000 [01:10<00:17, 19.05it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 66%|██████▌ | 662/1000 [01:10<00:17, 19.81it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 66%|██████▋ | 665/1000 [01:10<00:18, 18.52it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 67%|██████▋ | 668/1000 [01:10<00:17, 19.48it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 67%|██████▋ | 671/1000 [01:11<00:16, 20.13it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 67%|██████▋ | 673/1000 [01:11<00:17, 18.51it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 68%|██████▊ | 675/1000 [01:11<00:18, 17.34it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 68%|██████▊ | 678/1000 [01:11<00:17, 18.60it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 68%|██████▊ | 681/1000 [01:11<00:15, 20.54it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 68%|██████▊ | 684/1000 [01:11<00:14, 22.02it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 69%|██████▊ | 687/1000 [01:11<00:14, 21.95it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 69%|██████▉ | 690/1000 [01:12<00:14, 21.88it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 69%|██████▉ | 693/1000 [01:12<00:14, 21.89it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 70%|██████▉ | 696/1000 [01:12<00:15, 19.86it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 70%|██████▉ | 699/1000 [01:12<00:14, 20.38it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 70%|███████ | 702/1000 [01:12<00:14, 20.88it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 70%|███████ | 705/1000 [01:12<00:13, 22.25it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 71%|███████ | 708/1000 [01:12<00:13, 22.03it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 71%|███████ | 711/1000 [01:13<00:13, 21.96it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 71%|███████▏ | 714/1000 [01:13<00:13, 21.92it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 72%|███████▏ | 717/1000 [01:13<00:12, 21.95it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 72%|███████▏ | 720/1000 [01:13<00:12, 22.02it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 72%|███████▏ | 723/1000 [01:13<00:11, 23.15it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 73%|███████▎ | 726/1000 [01:13<00:11, 23.96it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 73%|███████▎ | 729/1000 [01:13<00:11, 23.29it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 73%|███████▎ | 731/1000 [01:13<00:13, 20.60it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 73%|███████▎ | 734/1000 [01:14<00:12, 21.03it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 74%|███████▎ | 737/1000 [01:14<00:11, 22.52it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 74%|███████▍ | 740/1000 [01:14<00:11, 22.26it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 74%|███████▍ | 743/1000 [01:14<00:12, 20.11it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 75%|███████▍ | 746/1000 [01:14<00:12, 20.59it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 75%|███████▍ | 749/1000 [01:14<00:11, 20.95it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 75%|███████▌ | 752/1000 [01:14<00:11, 20.98it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 75%|███████▌ | 754/1000 [01:15<00:13, 18.81it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 76%|███████▌ | 756/1000 [01:15<00:13, 17.50it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 76%|███████▌ | 759/1000 [01:15<00:12, 18.77it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 76%|███████▌ | 762/1000 [01:15<00:12, 19.62it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 76%|███████▋ | 765/1000 [01:15<00:11, 20.20it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 77%|███████▋ | 767/1000 [01:15<00:11, 19.47it/s, 31 steps of size 7.44e-02. acc. prob=0.92]
sample: 77%|███████▋ | 770/1000 [01:15<00:11, 20.21it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 77%|███████▋ | 773/1000 [01:15<00:10, 20.72it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 78%|███████▊ | 776/1000 [01:16<00:11, 19.11it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 78%|███████▊ | 779/1000 [01:16<00:11, 19.84it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 78%|███████▊ | 782/1000 [01:16<00:10, 20.36it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 78%|███████▊ | 785/1000 [01:16<00:10, 20.79it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 79%|███████▉ | 788/1000 [01:16<00:10, 21.11it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 79%|███████▉ | 791/1000 [01:16<00:09, 21.34it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 79%|███████▉ | 794/1000 [01:16<00:09, 21.52it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 80%|███████▉ | 797/1000 [01:17<00:09, 21.68it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 80%|████████ | 800/1000 [01:17<00:09, 21.65it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 80%|████████ | 803/1000 [01:17<00:09, 21.75it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 80%|████████ | 805/1000 [01:17<00:09, 19.64it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 81%|████████ | 808/1000 [01:17<00:10, 18.40it/s, 127 steps of size 7.44e-02. acc. prob=0.93]
sample: 81%|████████ | 811/1000 [01:17<00:09, 20.31it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 81%|████████▏ | 813/1000 [01:17<00:10, 18.62it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 82%|████████▏ | 816/1000 [01:18<00:09, 19.55it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 82%|████████▏ | 819/1000 [01:18<00:08, 20.21it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 82%|████████▏ | 822/1000 [01:18<00:08, 20.74it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 82%|████████▏ | 824/1000 [01:18<00:09, 18.94it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 83%|████████▎ | 826/1000 [01:18<00:09, 17.63it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 83%|████████▎ | 829/1000 [01:18<00:09, 18.85it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 83%|████████▎ | 832/1000 [01:18<00:08, 19.70it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 84%|████████▎ | 835/1000 [01:19<00:08, 20.39it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 84%|████████▍ | 838/1000 [01:19<00:07, 20.80it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 84%|████████▍ | 841/1000 [01:19<00:07, 21.13it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 84%|████████▍ | 844/1000 [01:19<00:07, 21.35it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 85%|████████▍ | 847/1000 [01:19<00:06, 22.65it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 85%|████████▌ | 850/1000 [01:19<00:06, 22.28it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 85%|████████▌ | 853/1000 [01:19<00:06, 22.05it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 86%|████████▌ | 856/1000 [01:20<00:06, 21.99it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 86%|████████▌ | 858/1000 [01:20<00:07, 17.93it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 86%|████████▌ | 861/1000 [01:20<00:07, 19.02it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 86%|████████▋ | 864/1000 [01:20<00:07, 18.90it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 87%|████████▋ | 867/1000 [01:20<00:06, 19.75it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 87%|████████▋ | 870/1000 [01:20<00:06, 21.36it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 87%|████████▋ | 873/1000 [01:20<00:05, 22.66it/s, 31 steps of size 7.44e-02. acc. prob=0.92]
sample: 88%|████████▊ | 876/1000 [01:21<00:05, 22.39it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 88%|████████▊ | 879/1000 [01:21<00:05, 22.21it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 88%|████████▊ | 882/1000 [01:21<00:05, 21.99it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 88%|████████▊ | 885/1000 [01:21<00:04, 23.12it/s, 31 steps of size 7.44e-02. acc. prob=0.92]
sample: 89%|████████▉ | 888/1000 [01:21<00:04, 22.71it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 89%|████████▉ | 891/1000 [01:21<00:04, 22.49it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 89%|████████▉ | 893/1000 [01:21<00:05, 18.97it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 90%|████████▉ | 896/1000 [01:21<00:05, 19.80it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 90%|████████▉ | 899/1000 [01:22<00:04, 20.88it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 90%|█████████ | 902/1000 [01:22<00:04, 21.95it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 91%|█████████ | 906/1000 [01:22<00:03, 25.14it/s, 31 steps of size 7.44e-02. acc. prob=0.92]
sample: 91%|█████████ | 910/1000 [01:22<00:03, 26.40it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 91%|█████████▏| 913/1000 [01:22<00:03, 26.17it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 92%|█████████▏| 916/1000 [01:22<00:03, 25.04it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 92%|█████████▏| 918/1000 [01:22<00:03, 21.75it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 92%|█████████▏| 921/1000 [01:23<00:03, 20.62it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 92%|█████████▏| 924/1000 [01:23<00:03, 22.03it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 93%|█████████▎| 927/1000 [01:23<00:03, 23.12it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 93%|█████████▎| 930/1000 [01:23<00:02, 23.99it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 93%|█████████▎| 933/1000 [01:23<00:02, 24.34it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 94%|█████████▎| 936/1000 [01:23<00:02, 24.59it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 94%|█████████▍| 939/1000 [01:23<00:02, 24.92it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 94%|█████████▍| 942/1000 [01:23<00:02, 25.01it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 95%|█████████▍| 946/1000 [01:23<00:02, 26.40it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 95%|█████████▍| 948/1000 [01:24<00:02, 23.69it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 95%|█████████▌| 950/1000 [01:24<00:02, 20.75it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 95%|█████████▌| 953/1000 [01:24<00:02, 22.12it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 96%|█████████▌| 956/1000 [01:24<00:01, 23.10it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 96%|█████████▌| 959/1000 [01:24<00:01, 24.10it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 96%|█████████▌| 962/1000 [01:24<00:01, 24.26it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 96%|█████████▋| 965/1000 [01:24<00:01, 23.19it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 97%|█████████▋| 968/1000 [01:25<00:01, 21.68it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 97%|█████████▋| 971/1000 [01:25<00:01, 22.64it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 97%|█████████▋| 973/1000 [01:25<00:01, 20.50it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 98%|█████████▊| 976/1000 [01:25<00:01, 20.92it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 98%|█████████▊| 979/1000 [01:25<00:00, 21.26it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 98%|█████████▊| 981/1000 [01:25<00:00, 19.24it/s, 127 steps of size 7.44e-02. acc. prob=0.93]
sample: 98%|█████████▊| 984/1000 [01:25<00:00, 20.11it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 99%|█████████▊| 987/1000 [01:25<00:00, 20.66it/s, 63 steps of size 7.44e-02. acc. prob=0.93]
sample: 99%|█████████▉| 990/1000 [01:26<00:00, 21.07it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 99%|█████████▉| 993/1000 [01:26<00:00, 21.36it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 100%|█████████▉| 996/1000 [01:26<00:00, 21.51it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
sample: 100%|█████████▉| 998/1000 [01:26<00:00, 19.43it/s, 127 steps of size 7.44e-02. acc. prob=0.92]
sample: 100%|██████████| 1000/1000 [01:26<00:00, 11.55it/s, 63 steps of size 7.44e-02. acc. prob=0.92]
We recommend running at least 4 chains for robust computation of convergence diagnostics
In [30]:
Copied!
#### posterior statistics of non-centered model
az.summary(
samples_model_reg_v_ex4_A1, var_names=["~_offset", "~_id"], filter_vars="like"
) # var_names= ["~_offset"])
#### posterior statistics of non-centered model
az.summary(
samples_model_reg_v_ex4_A1, var_names=["~_offset", "~_id"], filter_vars="like"
) # var_names= ["~_offset"])
Out[30]:
| mean | sd | eti89_lb | eti89_ub | ess_bulk | ess_tail | r_hat | mcse_mean | mcse_sd | |
|---|---|---|---|---|---|---|---|---|---|
| v_zGPe | 0.376 | 0.053 | 0.3 | 0.45 | 650 | 723 | 1.00 | 0.0022 | 0.0025 |
| v_C(diagnosis)[PD] | -0.237 | 0.061 | -0.33 | -0.14 | 1281 | 1178 | 1.00 | 0.0017 | 0.0012 |
| z | 0.4984 | 0.0062 | 0.49 | 0.51 | 2125 | 1195 | 1.00 | 0.00013 | 9.4e-05 |
| v_sevScore | -0.096 | 0.105 | -0.26 | 0.076 | 1197 | 1093 | 1.00 | 0.003 | 0.0023 |
| v_zSTN:sevScore:C(diagnosis)[PD] | -0.014 | 0.067 | -0.12 | 0.094 | 1222 | 1064 | 1.00 | 0.0019 | 0.0014 |
| v_sevScore:C(diagnosis)[PD] | -0.054 | 0.117 | -0.24 | 0.13 | 1096 | 1108 | 1.00 | 0.0035 | 0.0026 |
| v_zSTN:C(diagnosis)[PD] | 0.014 | 0.036 | -0.043 | 0.073 | 1243 | 992 | 1.00 | 0.001 | 0.00072 |
| v_zGPe:sevScore:C(diagnosis)[PD] | -0.003 | 0.058 | -0.098 | 0.087 | 1122 | 937 | 1.00 | 0.0017 | 0.0013 |
| v_zGPe:sevScore | -0.498 | 0.1 | -0.65 | -0.34 | 713 | 783 | 1.00 | 0.0038 | 0.0035 |
| v_zSTN | 0.856 | 0.066 | 0.75 | 0.95 | 567 | 640 | 1.00 | 0.0029 | 0.0028 |
| v_Intercept | 0.627 | 0.057 | 0.54 | 0.72 | 1288 | 1002 | 1.00 | 0.0016 | 0.0011 |
| v_zGPe:C(diagnosis)[PD] | -0.006 | 0.031 | -0.056 | 0.044 | 1189 | 1029 | 1.00 | 0.00091 | 0.00067 |
| v_zSTN:sevScore | -0.64 | 0.13 | -0.83 | -0.43 | 545 | 582 | 1.00 | 0.0058 | 0.0058 |
| t | 0.5182 | 0.0062 | 0.51 | 0.53 | 1700 | 1248 | 1.00 | 0.00015 | 0.00011 |
| a | 1.4978 | 0.0154 | 1.5 | 1.5 | 2226 | 1258 | 1.00 | 0.00033 | 0.00024 |
Other Example 2¶
In [31]:
Copied!
model_reg_v_ex4_A2 = hssm.HSSM(
data=combined_dataset3,
include=[
{
"name": "v",
"formula": (
"v ~ 1 + (zSTN + zGPe)*sevScore + "
"((1 + zSTN + zGPe)|participant_id/C(diagnosis))"
),
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zSTN:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
model_reg_v_ex4_A2 = hssm.HSSM(
data=combined_dataset3,
include=[
{
"name": "v",
"formula": (
"v ~ 1 + (zSTN + zGPe)*sevScore + "
"((1 + zSTN + zGPe)|participant_id/C(diagnosis))"
),
"prior": {
# All ways to specify priors in the non-regression case
# work the same way here.
"Intercept": {"name": "Normal", "mu": 1.5, "sigma": 1.0},
"zSTN": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe": {"name": "Normal", "mu": 0, "sigma": 1.0},
"sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zSTN:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"zGPe:sevScore": {"name": "Normal", "mu": 0, "sigma": 1.0},
"1|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zSTN|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
"zGPe|participant_id": {
"name": "Normal",
"mu": 0,
"sigma": {"name": "HalfNormal", "sigma": 1},
},
},
"link": "identity",
}
],
noncentered=True,
p_outlier=0.05,
)
Model initialized successfully.
In [32]:
Copied!
samples_model_reg_v_ex4_A2 = model_reg_v_ex4_A2.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
samples_model_reg_v_ex4_A2 = model_reg_v_ex4_A2.sample(
sampler="numpyro",
cores=3,
chains=3,
draws=500,
tune=500,
idata_kwargs=dict(log_likelihood=True),
)
Using default initvals.
NUTS[numpyro]: [a, z, t, v_Intercept, v_zSTN, v_zGPe, v_sevScore, v_zSTN:sevScore, v_zGPe:sevScore, v_1|participant_id_sigma, v_1|participant_id_offset, v_1|participant_id:C(diagnosis)_sigma, v_1|participant_id:C(diagnosis)_offset, v_zSTN|participant_id_sigma, v_zSTN|participant_id_offset, v_zSTN|participant_id:C(diagnosis)_mu, v_zSTN|participant_id:C(diagnosis)_sigma, v_zSTN|participant_id:C(diagnosis)_offset, v_zGPe|participant_id_sigma, v_zGPe|participant_id_offset, v_zGPe|participant_id:C(diagnosis)_mu, v_zGPe|participant_id:C(diagnosis)_sigma, v_zGPe|participant_id:C(diagnosis)_offset]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 0%| | 0/1000 [00:12<?, ?it/s, 1023 steps of size 2.53e-03. acc. prob=0.71]
warmup: 4%|▎ | 35/1000 [00:12<05:51, 2.75it/s, 1023 steps of size 4.14e-03. acc. prob=0.72]
warmup: 4%|▎ | 36/1000 [00:13<06:03, 2.65it/s, 1023 steps of size 5.69e-03. acc. prob=0.72]
warmup: 4%|▎ | 37/1000 [00:13<06:01, 2.66it/s, 511 steps of size 6.98e-03. acc. prob=0.72]
warmup: 4%|▍ | 38/1000 [00:14<05:59, 2.67it/s, 511 steps of size 7.26e-03. acc. prob=0.73]
warmup: 4%|▍ | 39/1000 [00:14<05:57, 2.69it/s, 511 steps of size 1.27e-02. acc. prob=0.73]
warmup: 4%|▍ | 41/1000 [00:15<05:51, 2.73it/s, 1023 steps of size 3.09e-03. acc. prob=0.72]
warmup: 4%|▍ | 42/1000 [00:15<06:23, 2.50it/s, 1023 steps of size 5.11e-03. acc. prob=0.73]
warmup: 4%|▍ | 43/1000 [00:16<06:49, 2.34it/s, 1023 steps of size 5.91e-03. acc. prob=0.73]
warmup: 4%|▍ | 44/1000 [00:16<06:23, 2.49it/s, 511 steps of size 3.17e-03. acc. prob=0.73]
warmup: 4%|▍ | 45/1000 [00:17<06:56, 2.30it/s, 1023 steps of size 5.28e-03. acc. prob=0.73]
warmup: 5%|▍ | 46/1000 [00:17<06:22, 2.49it/s, 511 steps of size 7.37e-03. acc. prob=0.74]
warmup: 5%|▍ | 47/1000 [00:17<05:54, 2.69it/s, 511 steps of size 1.05e-02. acc. prob=0.74]
warmup: 5%|▍ | 49/1000 [00:18<05:18, 2.98it/s, 1023 steps of size 2.69e-03. acc. prob=0.73]
warmup: 5%|▌ | 50/1000 [00:18<06:09, 2.57it/s, 1023 steps of size 4.55e-03. acc. prob=0.73]
warmup: 5%|▌ | 51/1000 [00:19<06:51, 2.31it/s, 1023 steps of size 7.17e-03. acc. prob=0.74]
warmup: 5%|▌ | 52/1000 [00:19<06:14, 2.53it/s, 511 steps of size 8.22e-03. acc. prob=0.74]
warmup: 5%|▌ | 53/1000 [00:20<05:45, 2.74it/s, 511 steps of size 5.79e-03. acc. prob=0.74]
warmup: 5%|▌ | 54/1000 [00:20<05:23, 2.93it/s, 511 steps of size 4.77e-03. acc. prob=0.74]
warmup: 6%|▌ | 55/1000 [00:20<05:06, 3.08it/s, 511 steps of size 4.81e-03. acc. prob=0.74]
warmup: 6%|▌ | 56/1000 [00:21<06:12, 2.54it/s, 1023 steps of size 7.64e-03. acc. prob=0.74]
warmup: 6%|▌ | 57/1000 [00:21<05:40, 2.77it/s, 511 steps of size 2.69e-03. acc. prob=0.74]
warmup: 6%|▌ | 58/1000 [00:22<06:38, 2.37it/s, 1023 steps of size 3.94e-03. acc. prob=0.74]
warmup: 6%|▌ | 59/1000 [00:22<07:23, 2.12it/s, 1023 steps of size 6.10e-03. acc. prob=0.74]
warmup: 6%|▌ | 60/1000 [00:22<06:32, 2.39it/s, 511 steps of size 9.93e-03. acc. prob=0.75]
warmup: 6%|▌ | 62/1000 [00:23<05:36, 2.79it/s, 1023 steps of size 3.02e-03. acc. prob=0.74]
warmup: 6%|▋ | 63/1000 [00:24<06:32, 2.39it/s, 1023 steps of size 4.87e-03. acc. prob=0.74]
warmup: 6%|▋ | 64/1000 [00:24<07:16, 2.14it/s, 1023 steps of size 6.81e-03. acc. prob=0.75]
warmup: 6%|▋ | 65/1000 [00:25<06:33, 2.38it/s, 511 steps of size 9.71e-03. acc. prob=0.75]
warmup: 7%|▋ | 67/1000 [00:25<05:42, 2.72it/s, 1023 steps of size 4.51e-03. acc. prob=0.75]
warmup: 7%|▋ | 68/1000 [00:25<05:27, 2.85it/s, 511 steps of size 5.13e-03. acc. prob=0.75]
warmup: 7%|▋ | 69/1000 [00:26<06:23, 2.43it/s, 1023 steps of size 7.74e-03. acc. prob=0.75]
warmup: 7%|▋ | 70/1000 [00:26<05:54, 2.62it/s, 511 steps of size 7.58e-03. acc. prob=0.75]
warmup: 7%|▋ | 71/1000 [00:27<05:33, 2.79it/s, 511 steps of size 2.75e-03. acc. prob=0.74]
warmup: 7%|▋ | 72/1000 [00:27<06:35, 2.35it/s, 1023 steps of size 4.26e-03. acc. prob=0.75]
warmup: 7%|▋ | 73/1000 [00:28<07:33, 2.04it/s, 1023 steps of size 6.17e-03. acc. prob=0.75]
warmup: 7%|▋ | 74/1000 [00:28<06:52, 2.24it/s, 511 steps of size 8.69e-03. acc. prob=0.75]
warmup: 8%|▊ | 75/1000 [00:29<06:27, 2.39it/s, 511 steps of size 3.61e-03. acc. prob=0.75]
warmup: 8%|▊ | 76/1000 [00:29<07:39, 2.01it/s, 1023 steps of size 5.27e-03. acc. prob=0.75]
warmup: 8%|▊ | 77/1000 [00:30<06:57, 2.21it/s, 511 steps of size 4.74e-03. acc. prob=0.75]
warmup: 8%|▊ | 78/1000 [00:30<08:03, 1.91it/s, 1023 steps of size 7.21e-03. acc. prob=0.75]
warmup: 8%|▊ | 79/1000 [00:31<07:15, 2.12it/s, 511 steps of size 1.10e-02. acc. prob=0.76]
warmup: 8%|▊ | 81/1000 [00:31<06:24, 2.39it/s, 1023 steps of size 3.93e-03. acc. prob=0.75]
warmup: 8%|▊ | 82/1000 [00:32<07:30, 2.04it/s, 1023 steps of size 5.71e-03. acc. prob=0.75]
warmup: 8%|▊ | 83/1000 [00:32<06:56, 2.20it/s, 511 steps of size 8.75e-03. acc. prob=0.76]
warmup: 8%|▊ | 84/1000 [00:33<06:30, 2.34it/s, 511 steps of size 2.44e-03. acc. prob=0.75]
warmup: 8%|▊ | 85/1000 [00:33<07:30, 2.03it/s, 1023 steps of size 3.72e-03. acc. prob=0.75]
warmup: 9%|▊ | 86/1000 [00:34<08:01, 1.90it/s, 1023 steps of size 5.67e-03. acc. prob=0.76]
warmup: 9%|▊ | 87/1000 [00:34<07:04, 2.15it/s, 511 steps of size 5.80e-03. acc. prob=0.76]
warmup: 9%|▉ | 88/1000 [00:35<06:19, 2.40it/s, 511 steps of size 8.43e-03. acc. prob=0.76]
warmup: 9%|▉ | 89/1000 [00:35<05:48, 2.61it/s, 511 steps of size 3.89e-03. acc. prob=0.75]
warmup: 9%|▉ | 90/1000 [00:36<06:51, 2.21it/s, 1023 steps of size 3.52e-03. acc. prob=0.75]
warmup: 9%|▉ | 91/1000 [00:36<07:33, 2.01it/s, 1023 steps of size 3.80e-03. acc. prob=0.75]
warmup: 9%|▉ | 92/1000 [00:37<07:58, 1.90it/s, 1023 steps of size 5.44e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:37<06:56, 2.18it/s, 511 steps of size 7.94e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:37<06:13, 2.43it/s, 511 steps of size 1.14e-02. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:38<05:25, 2.78it/s, 1023 steps of size 5.77e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:38<05:10, 2.91it/s, 511 steps of size 8.59e-03. acc. prob=0.76]
warmup: 10%|▉ | 98/1000 [00:39<05:03, 2.97it/s, 511 steps of size 2.46e-03. acc. prob=0.75]
warmup: 10%|▉ | 99/1000 [00:39<06:20, 2.37it/s, 1023 steps of size 3.61e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:40<07:28, 2.01it/s, 1023 steps of size 3.19e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:41<08:17, 1.81it/s, 1023 steps of size 4.59e-02. acc. prob=0.76]
warmup: 10%|█ | 103/1000 [00:41<04:59, 3.00it/s, 63 steps of size 1.55e-01. acc. prob=0.76]
warmup: 11%|█ | 106/1000 [00:41<03:00, 4.96it/s, 255 steps of size 4.64e-02. acc. prob=0.76]
warmup: 11%|█ | 108/1000 [00:41<02:20, 6.35it/s, 63 steps of size 1.33e-01. acc. prob=0.76]
warmup: 11%|█ | 112/1000 [00:41<01:24, 10.51it/s, 63 steps of size 9.80e-02. acc. prob=0.76]
warmup: 12%|█▏ | 115/1000 [00:41<01:06, 13.31it/s, 63 steps of size 1.76e-01. acc. prob=0.77]
warmup: 12%|█▏ | 118/1000 [00:41<01:03, 13.86it/s, 255 steps of size 4.05e-02. acc. prob=0.76]
warmup: 12%|█▏ | 120/1000 [00:42<01:02, 14.19it/s, 63 steps of size 1.22e-01. acc. prob=0.77]
warmup: 12%|█▏ | 124/1000 [00:42<00:58, 15.10it/s, 255 steps of size 4.15e-02. acc. prob=0.77]
warmup: 13%|█▎ | 126/1000 [00:42<00:57, 15.21it/s, 63 steps of size 1.25e-01. acc. prob=0.77]
warmup: 13%|█▎ | 131/1000 [00:42<00:52, 16.57it/s, 255 steps of size 3.97e-02. acc. prob=0.77]
warmup: 13%|█▎ | 133/1000 [00:42<00:56, 15.31it/s, 127 steps of size 1.15e-01. acc. prob=0.77]
warmup: 14%|█▎ | 137/1000 [00:43<00:49, 17.53it/s, 127 steps of size 5.39e-02. acc. prob=0.77]
warmup: 14%|█▍ | 139/1000 [00:43<00:50, 17.07it/s, 63 steps of size 1.22e-01. acc. prob=0.77]
warmup: 14%|█▍ | 142/1000 [00:43<00:45, 18.66it/s, 127 steps of size 7.92e-02. acc. prob=0.77]
warmup: 14%|█▍ | 145/1000 [00:43<00:42, 19.94it/s, 63 steps of size 1.13e-01. acc. prob=0.77]
warmup: 15%|█▍ | 149/1000 [00:43<00:40, 21.12it/s, 127 steps of size 5.33e-02. acc. prob=0.77]
warmup: 15%|█▌ | 151/1000 [00:43<00:43, 19.65it/s, 63 steps of size 1.12e+00. acc. prob=0.77]
warmup: 15%|█▌ | 154/1000 [00:43<00:39, 21.68it/s, 127 steps of size 4.88e-02. acc. prob=0.77]
warmup: 16%|█▌ | 156/1000 [00:43<00:42, 19.92it/s, 63 steps of size 1.29e-01. acc. prob=0.77]
warmup: 16%|█▌ | 161/1000 [00:44<00:31, 27.00it/s, 15 steps of size 2.14e-02. acc. prob=0.77]
warmup: 16%|█▌ | 162/1000 [00:44<00:43, 19.21it/s, 255 steps of size 4.05e-02. acc. prob=0.77]
warmup: 16%|█▋ | 164/1000 [00:44<00:45, 18.23it/s, 63 steps of size 1.33e-01. acc. prob=0.77]
warmup: 17%|█▋ | 167/1000 [00:44<00:40, 20.79it/s, 31 steps of size 5.66e-02. acc. prob=0.77]
warmup: 17%|█▋ | 170/1000 [00:44<00:36, 22.89it/s, 31 steps of size 6.09e-02. acc. prob=0.77]
warmup: 17%|█▋ | 172/1000 [00:44<00:38, 21.72it/s, 31 steps of size 1.83e-01. acc. prob=0.77]
warmup: 18%|█▊ | 175/1000 [00:44<00:34, 23.60it/s, 63 steps of size 1.32e-01. acc. prob=0.77]
warmup: 18%|█▊ | 180/1000 [00:44<00:27, 29.44it/s, 31 steps of size 2.18e-01. acc. prob=0.77]
warmup: 18%|█▊ | 183/1000 [00:45<00:30, 27.09it/s, 63 steps of size 1.19e-01. acc. prob=0.77]
warmup: 19%|█▊ | 187/1000 [00:45<00:28, 28.55it/s, 63 steps of size 1.68e-01. acc. prob=0.78]
warmup: 19%|█▉ | 189/1000 [00:45<00:30, 26.21it/s, 127 steps of size 6.68e-02. acc. prob=0.77]
warmup: 19%|█▉ | 192/1000 [00:45<00:29, 26.94it/s, 31 steps of size 2.59e-01. acc. prob=0.78]
warmup: 20%|█▉ | 196/1000 [00:45<00:36, 22.11it/s, 255 steps of size 4.19e-02. acc. prob=0.77]
warmup: 20%|█▉ | 198/1000 [00:45<00:38, 20.58it/s, 63 steps of size 9.37e-02. acc. prob=0.78]
warmup: 20%|██ | 202/1000 [00:45<00:33, 23.78it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 20%|██ | 205/1000 [00:45<00:31, 25.12it/s, 63 steps of size 4.27e-02. acc. prob=0.77]
warmup: 21%|██ | 207/1000 [00:46<00:35, 22.44it/s, 63 steps of size 1.04e-01. acc. prob=0.78]
warmup: 21%|██ | 211/1000 [00:46<00:34, 23.15it/s, 127 steps of size 7.88e-02. acc. prob=0.78]
warmup: 22%|██▏ | 215/1000 [00:46<00:31, 24.93it/s, 63 steps of size 1.03e-01. acc. prob=0.78]
warmup: 22%|██▏ | 219/1000 [00:46<00:28, 27.03it/s, 63 steps of size 1.13e-01. acc. prob=0.78]
warmup: 22%|██▏ | 223/1000 [00:46<00:25, 30.05it/s, 31 steps of size 3.26e-02. acc. prob=0.77]
warmup: 22%|██▎ | 225/1000 [00:46<00:32, 23.95it/s, 127 steps of size 6.46e-02. acc. prob=0.78]
warmup: 23%|██▎ | 228/1000 [00:46<00:30, 25.27it/s, 31 steps of size 1.45e-01. acc. prob=0.78]
warmup: 23%|██▎ | 233/1000 [00:46<00:24, 31.38it/s, 31 steps of size 1.31e-01. acc. prob=0.78]
warmup: 24%|██▍ | 238/1000 [00:47<00:21, 36.14it/s, 31 steps of size 8.49e-02. acc. prob=0.78]
warmup: 24%|██▍ | 241/1000 [00:47<00:23, 32.42it/s, 63 steps of size 1.46e-01. acc. prob=0.78]
warmup: 25%|██▍ | 247/1000 [00:47<00:20, 36.87it/s, 63 steps of size 9.45e-02. acc. prob=0.78]
warmup: 25%|██▌ | 250/1000 [00:47<00:21, 34.80it/s, 31 steps of size 1.57e-01. acc. prob=0.78]
warmup: 25%|██▌ | 253/1000 [00:47<00:22, 33.15it/s, 127 steps of size 4.06e-02. acc. prob=0.78]
warmup: 26%|██▌ | 255/1000 [00:47<00:29, 25.07it/s, 127 steps of size 8.24e-02. acc. prob=0.78]
warmup: 26%|██▌ | 258/1000 [00:47<00:29, 25.14it/s, 63 steps of size 1.64e-01. acc. prob=0.78]
warmup: 26%|██▌ | 262/1000 [00:47<00:26, 27.59it/s, 63 steps of size 1.28e-01. acc. prob=0.78]
warmup: 26%|██▋ | 265/1000 [00:48<00:27, 26.69it/s, 127 steps of size 6.89e-02. acc. prob=0.78]
warmup: 27%|██▋ | 268/1000 [00:48<00:26, 27.28it/s, 31 steps of size 6.75e-02. acc. prob=0.78]
warmup: 27%|██▋ | 271/1000 [00:48<00:26, 27.90it/s, 31 steps of size 7.32e-02. acc. prob=0.78]
warmup: 28%|██▊ | 275/1000 [00:48<00:31, 22.99it/s, 255 steps of size 3.62e-02. acc. prob=0.78]
warmup: 28%|██▊ | 277/1000 [00:48<00:34, 21.11it/s, 63 steps of size 3.51e-02. acc. prob=0.78]
warmup: 28%|██▊ | 279/1000 [00:48<00:36, 19.74it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 28%|██▊ | 282/1000 [00:48<00:32, 22.08it/s, 63 steps of size 1.32e-01. acc. prob=0.78]
warmup: 28%|██▊ | 285/1000 [00:48<00:29, 23.95it/s, 63 steps of size 1.23e-01. acc. prob=0.78]
warmup: 29%|██▉ | 288/1000 [00:49<00:28, 25.33it/s, 63 steps of size 1.62e-01. acc. prob=0.78]
warmup: 29%|██▉ | 293/1000 [00:49<00:26, 26.28it/s, 127 steps of size 6.04e-02. acc. prob=0.78]
warmup: 30%|██▉ | 296/1000 [00:49<00:26, 27.01it/s, 31 steps of size 1.67e-01. acc. prob=0.78]
warmup: 30%|██▉ | 298/1000 [00:49<00:28, 25.02it/s, 127 steps of size 5.41e-02. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:49<00:31, 22.33it/s, 63 steps of size 1.17e-01. acc. prob=0.78]
warmup: 30%|███ | 304/1000 [00:49<00:25, 26.85it/s, 31 steps of size 3.42e-02. acc. prob=0.78]
warmup: 31%|███ | 306/1000 [00:49<00:29, 23.62it/s, 63 steps of size 8.46e-02. acc. prob=0.78]
warmup: 31%|███ | 311/1000 [00:49<00:24, 27.93it/s, 63 steps of size 9.38e-02. acc. prob=0.78]
warmup: 31%|███▏ | 314/1000 [00:50<00:25, 26.83it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 32%|███▏ | 317/1000 [00:50<00:27, 24.78it/s, 127 steps of size 6.83e-02. acc. prob=0.78]
warmup: 32%|███▏ | 320/1000 [00:50<00:27, 24.67it/s, 63 steps of size 8.06e-02. acc. prob=0.78]
warmup: 32%|███▎ | 325/1000 [00:50<00:23, 29.13it/s, 63 steps of size 7.48e-02. acc. prob=0.78]
warmup: 33%|███▎ | 329/1000 [00:50<00:21, 31.86it/s, 31 steps of size 3.58e-02. acc. prob=0.78]
warmup: 33%|███▎ | 331/1000 [00:50<00:26, 24.90it/s, 127 steps of size 7.72e-02. acc. prob=0.78]
warmup: 34%|███▎ | 335/1000 [00:50<00:23, 28.61it/s, 31 steps of size 3.77e-02. acc. prob=0.78]
warmup: 34%|███▎ | 337/1000 [00:50<00:26, 24.86it/s, 63 steps of size 8.21e-02. acc. prob=0.78]
warmup: 34%|███▍ | 340/1000 [00:51<00:29, 22.38it/s, 127 steps of size 6.77e-02. acc. prob=0.78]
warmup: 34%|███▍ | 343/1000 [00:51<00:27, 24.07it/s, 31 steps of size 8.61e-02. acc. prob=0.78]
warmup: 35%|███▍ | 347/1000 [00:51<00:23, 28.00it/s, 31 steps of size 6.96e-02. acc. prob=0.78]
warmup: 35%|███▌ | 351/1000 [00:51<00:21, 29.92it/s, 31 steps of size 4.25e-02. acc. prob=0.78]
warmup: 35%|███▌ | 353/1000 [00:51<00:25, 25.69it/s, 63 steps of size 8.67e-02. acc. prob=0.78]
warmup: 36%|███▌ | 357/1000 [00:51<00:23, 27.74it/s, 63 steps of size 9.59e-02. acc. prob=0.78]
warmup: 36%|███▌ | 360/1000 [00:51<00:22, 28.14it/s, 63 steps of size 8.01e-02. acc. prob=0.78]
warmup: 36%|███▋ | 363/1000 [00:51<00:22, 28.44it/s, 31 steps of size 1.20e-01. acc. prob=0.78]
warmup: 37%|███▋ | 366/1000 [00:51<00:22, 28.67it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 37%|███▋ | 370/1000 [00:52<00:20, 31.48it/s, 31 steps of size 1.32e-01. acc. prob=0.78]
warmup: 37%|███▋ | 372/1000 [00:52<00:22, 27.89it/s, 127 steps of size 6.63e-02. acc. prob=0.78]
warmup: 38%|███▊ | 375/1000 [00:52<00:22, 28.26it/s, 31 steps of size 5.00e-02. acc. prob=0.78]
warmup: 38%|███▊ | 377/1000 [00:52<00:25, 24.21it/s, 63 steps of size 7.79e-02. acc. prob=0.78]
warmup: 38%|███▊ | 380/1000 [00:52<00:25, 24.29it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 38%|███▊ | 384/1000 [00:52<00:21, 28.44it/s, 31 steps of size 1.44e-01. acc. prob=0.78]
warmup: 39%|███▉ | 389/1000 [00:52<00:17, 34.36it/s, 31 steps of size 8.74e-02. acc. prob=0.78]
warmup: 39%|███▉ | 393/1000 [00:52<00:17, 35.41it/s, 9 steps of size 5.96e-02. acc. prob=0.78]
warmup: 40%|███▉ | 395/1000 [00:52<00:20, 29.27it/s, 63 steps of size 9.82e-02. acc. prob=0.78]
warmup: 40%|███▉ | 399/1000 [00:53<00:18, 32.11it/s, 31 steps of size 1.52e-01. acc. prob=0.78]
warmup: 40%|████ | 402/1000 [00:53<00:19, 31.21it/s, 63 steps of size 1.01e-01. acc. prob=0.78]
warmup: 41%|████ | 406/1000 [00:53<00:18, 31.59it/s, 63 steps of size 1.11e-01. acc. prob=0.78]
warmup: 41%|████ | 410/1000 [00:53<00:17, 33.60it/s, 31 steps of size 1.12e-01. acc. prob=0.78]
warmup: 42%|████▏ | 415/1000 [00:53<00:16, 35.76it/s, 63 steps of size 8.95e-02. acc. prob=0.78]
warmup: 42%|████▏ | 419/1000 [00:53<00:16, 34.68it/s, 63 steps of size 1.19e-01. acc. prob=0.78]
warmup: 42%|████▏ | 424/1000 [00:53<00:16, 35.39it/s, 63 steps of size 7.38e-02. acc. prob=0.78]
warmup: 43%|████▎ | 427/1000 [00:53<00:16, 33.75it/s, 31 steps of size 1.47e-01. acc. prob=0.78]
warmup: 43%|████▎ | 430/1000 [00:53<00:17, 32.43it/s, 63 steps of size 1.22e-01. acc. prob=0.78]
warmup: 43%|████▎ | 434/1000 [00:54<00:16, 34.10it/s, 31 steps of size 1.54e-01. acc. prob=0.78]
warmup: 44%|████▍ | 438/1000 [00:54<00:16, 33.41it/s, 63 steps of size 1.26e-01. acc. prob=0.78]
warmup: 44%|████▍ | 442/1000 [00:54<00:15, 35.00it/s, 63 steps of size 1.43e-01. acc. prob=0.78]
warmup: 45%|████▍ | 446/1000 [00:54<00:15, 35.93it/s, 63 steps of size 1.22e-01. acc. prob=0.78]
warmup: 45%|████▌ | 450/1000 [00:54<00:14, 36.78it/s, 31 steps of size 4.54e-02. acc. prob=0.78]
warmup: 45%|████▌ | 453/1000 [00:54<00:15, 34.38it/s, 31 steps of size 1.45e-01. acc. prob=0.78]
warmup: 46%|████▌ | 455/1000 [00:54<00:18, 30.00it/s, 127 steps of size 4.44e-02. acc. prob=0.78]
warmup: 46%|████▌ | 457/1000 [00:54<00:21, 25.30it/s, 63 steps of size 8.62e-02. acc. prob=0.78]
warmup: 46%|████▌ | 460/1000 [00:54<00:20, 26.35it/s, 31 steps of size 3.85e-02. acc. prob=0.78]
warmup: 46%|████▌ | 462/1000 [00:55<00:23, 23.01it/s, 63 steps of size 1.31e-01. acc. prob=0.78]
warmup: 46%|████▋ | 465/1000 [00:55<00:22, 23.42it/s, 127 steps of size 7.76e-02. acc. prob=0.78]
warmup: 47%|████▋ | 469/1000 [00:55<00:19, 27.74it/s, 31 steps of size 1.25e-01. acc. prob=0.78]
warmup: 47%|████▋ | 473/1000 [00:55<00:17, 29.39it/s, 63 steps of size 8.33e-02. acc. prob=0.78]
warmup: 48%|████▊ | 477/1000 [00:55<00:16, 32.25it/s, 31 steps of size 1.93e-01. acc. prob=0.78]
warmup: 48%|████▊ | 479/1000 [00:55<00:21, 24.02it/s, 255 steps of size 3.73e-02. acc. prob=0.78]
warmup: 48%|████▊ | 481/1000 [00:55<00:23, 21.64it/s, 63 steps of size 1.01e-01. acc. prob=0.78]
warmup: 48%|████▊ | 484/1000 [00:55<00:23, 22.40it/s, 63 steps of size 4.92e-02. acc. prob=0.78]
warmup: 49%|████▊ | 486/1000 [00:56<00:24, 20.60it/s, 63 steps of size 7.06e-02. acc. prob=0.78]
warmup: 49%|████▉ | 490/1000 [00:56<00:22, 22.80it/s, 127 steps of size 6.64e-02. acc. prob=0.78]
warmup: 50%|████▉ | 495/1000 [00:56<00:19, 25.30it/s, 127 steps of size 5.20e-02. acc. prob=0.78]
warmup: 50%|████▉ | 497/1000 [00:56<00:21, 22.97it/s, 63 steps of size 4.43e-02. acc. prob=0.78]
warmup: 50%|████▉ | 499/1000 [00:56<00:23, 21.06it/s, 63 steps of size 9.27e-02. acc. prob=0.78]
sample: 50%|█████ | 502/1000 [00:56<00:22, 21.97it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 50%|█████ | 505/1000 [00:56<00:21, 22.63it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 51%|█████ | 508/1000 [00:56<00:21, 23.12it/s, 63 steps of size 8.04e-02. acc. prob=0.93]
sample: 51%|█████ | 511/1000 [00:57<00:21, 22.36it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 51%|█████▏ | 514/1000 [00:57<00:21, 22.49it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 52%|█████▏ | 517/1000 [00:57<00:21, 22.93it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 52%|█████▏ | 520/1000 [00:57<00:20, 23.16it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 52%|█████▏ | 523/1000 [00:57<00:21, 22.66it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 52%|█████▎ | 525/1000 [00:57<00:22, 20.86it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 53%|█████▎ | 527/1000 [00:57<00:23, 19.75it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 53%|█████▎ | 530/1000 [00:58<00:22, 20.64it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 53%|█████▎ | 533/1000 [00:58<00:21, 21.52it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 54%|█████▎ | 536/1000 [00:58<00:21, 21.61it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 54%|█████▍ | 539/1000 [00:58<00:20, 22.06it/s, 63 steps of size 8.04e-02. acc. prob=0.93]
sample: 54%|█████▍ | 542/1000 [00:58<00:20, 22.57it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 55%|█████▍ | 545/1000 [00:58<00:19, 22.93it/s, 63 steps of size 8.04e-02. acc. prob=0.93]
sample: 55%|█████▍ | 548/1000 [00:58<00:19, 23.28it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 55%|█████▌ | 551/1000 [00:58<00:19, 23.56it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 55%|█████▌ | 554/1000 [00:59<00:18, 23.61it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 56%|█████▌ | 557/1000 [00:59<00:17, 24.90it/s, 31 steps of size 8.04e-02. acc. prob=0.92]
sample: 56%|█████▌ | 560/1000 [00:59<00:17, 24.68it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 56%|█████▋ | 563/1000 [00:59<00:17, 24.55it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 57%|█████▋ | 566/1000 [00:59<00:17, 24.45it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 57%|█████▋ | 569/1000 [00:59<00:17, 24.39it/s, 63 steps of size 8.04e-02. acc. prob=0.92]
sample: 57%|█████▋ | 572/1000 [00:59<00:17, 24.34it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 57%|█████▊ | 575/1000 [00:59<00:17, 24.28it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 58%|█████▊ | 578/1000 [01:00<00:16, 25.69it/s, 31 steps of size 8.04e-02. acc. prob=0.91]
sample: 58%|█████▊ | 581/1000 [01:00<00:16, 25.57it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 58%|█████▊ | 584/1000 [01:00<00:16, 25.11it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 59%|█████▊ | 587/1000 [01:00<00:16, 24.83it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 59%|█████▉ | 590/1000 [01:00<00:16, 24.62it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 59%|█████▉ | 593/1000 [01:00<00:16, 24.14it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 60%|█████▉ | 596/1000 [01:00<00:16, 24.26it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 60%|█████▉ | 599/1000 [01:00<00:16, 24.22it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 60%|██████ | 602/1000 [01:00<00:15, 25.43it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 60%|██████ | 605/1000 [01:01<00:15, 24.87it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 61%|██████ | 608/1000 [01:01<00:15, 24.55it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 61%|██████ | 611/1000 [01:01<00:15, 24.56it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 61%|██████▏ | 614/1000 [01:01<00:15, 25.15it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 62%|██████▏ | 617/1000 [01:01<00:14, 26.05it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 62%|██████▏ | 620/1000 [01:01<00:14, 26.76it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 62%|██████▏ | 623/1000 [01:01<00:14, 26.44it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 63%|██████▎ | 626/1000 [01:01<00:14, 25.98it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 63%|██████▎ | 629/1000 [01:02<00:14, 25.89it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 63%|██████▎ | 632/1000 [01:02<00:14, 25.93it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 64%|██████▎ | 635/1000 [01:02<00:14, 25.50it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 64%|██████▍ | 638/1000 [01:02<00:14, 25.06it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 64%|██████▍ | 641/1000 [01:02<00:14, 25.19it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 64%|██████▍ | 644/1000 [01:02<00:14, 25.42it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 65%|██████▍ | 647/1000 [01:02<00:14, 25.13it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 65%|██████▌ | 650/1000 [01:02<00:13, 25.02it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 65%|██████▌ | 653/1000 [01:02<00:13, 24.89it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 66%|██████▌ | 656/1000 [01:03<00:13, 25.10it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 66%|██████▌ | 659/1000 [01:03<00:13, 25.53it/s, 63 steps of size 8.04e-02. acc. prob=0.91]
sample: 66%|██████▋ | 663/1000 [01:03<00:12, 26.64it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 67%|██████▋ | 666/1000 [01:03<00:12, 26.11it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 67%|██████▋ | 669/1000 [01:03<00:12, 25.73it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 67%|██████▋ | 672/1000 [01:03<00:12, 25.40it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 68%|██████▊ | 675/1000 [01:03<00:12, 25.34it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 68%|██████▊ | 678/1000 [01:03<00:12, 25.80it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 68%|██████▊ | 681/1000 [01:04<00:12, 26.48it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 68%|██████▊ | 684/1000 [01:04<00:11, 26.97it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 69%|██████▊ | 687/1000 [01:04<00:11, 26.76it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 69%|██████▉ | 690/1000 [01:04<00:11, 26.57it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 69%|██████▉ | 693/1000 [01:04<00:11, 26.42it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 70%|██████▉ | 696/1000 [01:04<00:11, 26.27it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 70%|██████▉ | 699/1000 [01:04<00:11, 26.62it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 70%|███████ | 702/1000 [01:04<00:11, 26.50it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 70%|███████ | 705/1000 [01:04<00:11, 26.50it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 71%|███████ | 708/1000 [01:05<00:11, 26.47it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 71%|███████ | 711/1000 [01:05<00:11, 26.16it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 71%|███████▏ | 714/1000 [01:05<00:10, 26.09it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 72%|███████▏ | 717/1000 [01:05<00:10, 26.10it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 72%|███████▏ | 720/1000 [01:05<00:10, 26.13it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 72%|███████▏ | 723/1000 [01:05<00:10, 26.06it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 73%|███████▎ | 726/1000 [01:05<00:10, 26.18it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 73%|███████▎ | 729/1000 [01:05<00:10, 26.13it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 73%|███████▎ | 732/1000 [01:05<00:10, 26.14it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 74%|███████▎ | 735/1000 [01:06<00:10, 26.14it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 74%|███████▍ | 738/1000 [01:06<00:10, 26.05it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 74%|███████▍ | 741/1000 [01:06<00:09, 25.94it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 74%|███████▍ | 744/1000 [01:06<00:09, 25.96it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 75%|███████▍ | 747/1000 [01:06<00:09, 25.99it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 75%|███████▌ | 750/1000 [01:06<00:09, 25.91it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 75%|███████▌ | 753/1000 [01:06<00:09, 26.05it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 76%|███████▌ | 756/1000 [01:06<00:09, 25.98it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 76%|███████▌ | 759/1000 [01:07<00:09, 26.08it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 76%|███████▌ | 762/1000 [01:07<00:09, 26.08it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 77%|███████▋ | 766/1000 [01:07<00:08, 27.33it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 77%|███████▋ | 769/1000 [01:07<00:08, 27.01it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 77%|███████▋ | 772/1000 [01:07<00:08, 26.80it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 78%|███████▊ | 775/1000 [01:07<00:08, 26.56it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 78%|███████▊ | 779/1000 [01:07<00:07, 27.67it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 78%|███████▊ | 782/1000 [01:07<00:08, 27.18it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 78%|███████▊ | 785/1000 [01:07<00:07, 26.92it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 79%|███████▉ | 788/1000 [01:08<00:07, 26.76it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 79%|███████▉ | 791/1000 [01:08<00:07, 26.96it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 79%|███████▉ | 794/1000 [01:08<00:07, 27.42it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 80%|███████▉ | 797/1000 [01:08<00:07, 27.79it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 80%|████████ | 800/1000 [01:08<00:07, 27.92it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 80%|████████ | 802/1000 [01:08<00:07, 24.98it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 80%|████████ | 804/1000 [01:08<00:08, 22.89it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 81%|████████ | 807/1000 [01:08<00:07, 24.29it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 81%|████████ | 810/1000 [01:08<00:07, 25.30it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 81%|████████▏ | 813/1000 [01:09<00:07, 26.16it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 82%|████████▏ | 816/1000 [01:09<00:06, 26.61it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 82%|████████▏ | 819/1000 [01:09<00:06, 26.87it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 82%|████████▏ | 822/1000 [01:09<00:06, 27.09it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 82%|████████▎ | 825/1000 [01:09<00:06, 27.11it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 83%|████████▎ | 828/1000 [01:09<00:06, 27.13it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 83%|████████▎ | 831/1000 [01:09<00:06, 27.12it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 83%|████████▎ | 834/1000 [01:09<00:06, 27.09it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 84%|████████▎ | 837/1000 [01:09<00:05, 27.18it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 84%|████████▍ | 840/1000 [01:10<00:05, 27.23it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 84%|████████▍ | 843/1000 [01:10<00:05, 27.26it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 85%|████████▍ | 846/1000 [01:10<00:05, 27.13it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 85%|████████▍ | 849/1000 [01:10<00:05, 27.17it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 85%|████████▌ | 852/1000 [01:10<00:05, 27.14it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 86%|████████▌ | 855/1000 [01:10<00:05, 27.14it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 86%|████████▌ | 858/1000 [01:10<00:05, 27.16it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 86%|████████▌ | 861/1000 [01:10<00:05, 27.18it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 86%|████████▋ | 864/1000 [01:10<00:05, 27.16it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 87%|████████▋ | 867/1000 [01:11<00:04, 27.16it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 87%|████████▋ | 870/1000 [01:11<00:04, 27.15it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 87%|████████▋ | 873/1000 [01:11<00:04, 27.21it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 88%|████████▊ | 876/1000 [01:11<00:04, 27.42it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 88%|████████▊ | 879/1000 [01:11<00:04, 27.29it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 88%|████████▊ | 882/1000 [01:11<00:04, 27.32it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 88%|████████▊ | 885/1000 [01:11<00:04, 27.36it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 89%|████████▉ | 888/1000 [01:11<00:04, 27.36it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 89%|████████▉ | 891/1000 [01:11<00:03, 27.43it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 89%|████████▉ | 894/1000 [01:12<00:03, 27.36it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 90%|████████▉ | 897/1000 [01:12<00:03, 27.34it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 90%|█████████ | 900/1000 [01:12<00:03, 27.27it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 90%|█████████ | 903/1000 [01:12<00:03, 27.39it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 91%|█████████ | 906/1000 [01:12<00:03, 27.46it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 91%|█████████ | 909/1000 [01:12<00:03, 27.37it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 91%|█████████ | 912/1000 [01:12<00:03, 27.34it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 92%|█████████▏| 915/1000 [01:12<00:03, 27.37it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 92%|█████████▏| 919/1000 [01:12<00:02, 28.72it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 92%|█████████▏| 922/1000 [01:13<00:02, 28.29it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 92%|█████████▎| 925/1000 [01:13<00:02, 28.05it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 93%|█████████▎| 928/1000 [01:13<00:02, 27.81it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 93%|█████████▎| 931/1000 [01:13<00:02, 27.66it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 93%|█████████▎| 934/1000 [01:13<00:02, 27.62it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 94%|█████████▎| 937/1000 [01:13<00:02, 27.58it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 94%|█████████▍| 940/1000 [01:13<00:02, 27.77it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 94%|█████████▍| 943/1000 [01:13<00:02, 27.93it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 95%|█████████▍| 946/1000 [01:13<00:01, 28.10it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 95%|█████████▍| 949/1000 [01:14<00:01, 28.22it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 95%|█████████▌| 952/1000 [01:14<00:01, 28.14it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 96%|█████████▌| 955/1000 [01:14<00:01, 28.03it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 96%|█████████▌| 959/1000 [01:14<00:01, 29.39it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 96%|█████████▌| 962/1000 [01:14<00:01, 29.00it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 96%|█████████▋| 965/1000 [01:14<00:01, 28.69it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 97%|█████████▋| 968/1000 [01:14<00:01, 28.53it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 97%|█████████▋| 971/1000 [01:14<00:01, 28.38it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 97%|█████████▋| 974/1000 [01:14<00:00, 28.37it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 98%|█████████▊| 977/1000 [01:14<00:00, 28.39it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 98%|█████████▊| 980/1000 [01:15<00:00, 28.33it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 98%|█████████▊| 983/1000 [01:15<00:00, 26.00it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 99%|█████████▊| 986/1000 [01:15<00:00, 26.50it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 99%|█████████▉| 989/1000 [01:15<00:00, 26.74it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 99%|█████████▉| 992/1000 [01:15<00:00, 27.07it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 100%|█████████▉| 995/1000 [01:15<00:00, 27.43it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 100%|█████████▉| 998/1000 [01:15<00:00, 27.61it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
sample: 100%|██████████| 1000/1000 [01:15<00:00, 13.18it/s, 63 steps of size 8.04e-02. acc. prob=0.90]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 0%| | 0/1000 [00:16<?, ?it/s, 1023 steps of size 2.05e-03. acc. prob=0.72]
warmup: 5%|▍ | 46/1000 [00:17<05:58, 2.66it/s, 1023 steps of size 3.37e-03. acc. prob=0.73]
warmup: 5%|▍ | 47/1000 [00:18<06:07, 2.59it/s, 1023 steps of size 4.47e-03. acc. prob=0.73]
warmup: 5%|▍ | 48/1000 [00:18<06:20, 2.50it/s, 1023 steps of size 7.11e-03. acc. prob=0.74]
warmup: 5%|▍ | 49/1000 [00:19<06:18, 2.51it/s, 511 steps of size 8.61e-03. acc. prob=0.74]
warmup: 5%|▌ | 50/1000 [00:19<05:58, 2.65it/s, 181 steps of size 1.44e-03. acc. prob=0.72]
warmup: 5%|▌ | 51/1000 [00:19<06:28, 2.44it/s, 1023 steps of size 2.41e-03. acc. prob=0.73]
warmup: 5%|▌ | 52/1000 [00:20<07:02, 2.24it/s, 1023 steps of size 4.04e-03. acc. prob=0.74]
warmup: 5%|▌ | 53/1000 [00:21<07:39, 2.06it/s, 1023 steps of size 5.73e-03. acc. prob=0.74]
warmup: 5%|▌ | 54/1000 [00:21<07:18, 2.16it/s, 511 steps of size 9.37e-03. acc. prob=0.74]
warmup: 6%|▌ | 56/1000 [00:22<06:46, 2.32it/s, 1023 steps of size 2.62e-03. acc. prob=0.74]
warmup: 6%|▌ | 57/1000 [00:23<07:36, 2.06it/s, 1023 steps of size 4.16e-03. acc. prob=0.74]
warmup: 6%|▌ | 58/1000 [00:23<08:22, 1.88it/s, 1023 steps of size 5.63e-03. acc. prob=0.74]
warmup: 6%|▌ | 59/1000 [00:24<07:41, 2.04it/s, 511 steps of size 7.39e-03. acc. prob=0.75]
warmup: 6%|▌ | 60/1000 [00:24<07:08, 2.19it/s, 511 steps of size 1.02e-02. acc. prob=0.75]
warmup: 6%|▌ | 62/1000 [00:25<06:28, 2.42it/s, 1023 steps of size 4.87e-03. acc. prob=0.74]
warmup: 6%|▋ | 63/1000 [00:25<07:33, 2.07it/s, 1023 steps of size 5.65e-03. acc. prob=0.75]
warmup: 6%|▋ | 64/1000 [00:26<07:04, 2.20it/s, 511 steps of size 8.90e-03. acc. prob=0.75]
warmup: 7%|▋ | 66/1000 [00:27<06:48, 2.29it/s, 1023 steps of size 3.60e-03. acc. prob=0.74]
warmup: 7%|▋ | 67/1000 [00:27<07:48, 1.99it/s, 1023 steps of size 5.28e-03. acc. prob=0.75]
warmup: 7%|▋ | 68/1000 [00:28<08:33, 1.81it/s, 1023 steps of size 8.30e-03. acc. prob=0.75]
warmup: 7%|▋ | 69/1000 [00:28<07:46, 2.00it/s, 511 steps of size 5.61e-03. acc. prob=0.75]
warmup: 7%|▋ | 70/1000 [00:29<07:08, 2.17it/s, 511 steps of size 8.81e-03. acc. prob=0.75]
warmup: 7%|▋ | 71/1000 [00:29<06:00, 2.58it/s, 282 steps of size 1.78e-03. acc. prob=0.74]
warmup: 7%|▋ | 72/1000 [00:30<07:27, 2.07it/s, 1023 steps of size 2.80e-03. acc. prob=0.75]
warmup: 7%|▋ | 73/1000 [00:30<08:29, 1.82it/s, 1023 steps of size 4.06e-03. acc. prob=0.75]
warmup: 7%|▋ | 74/1000 [00:31<09:20, 1.65it/s, 1023 steps of size 3.38e-03. acc. prob=0.75]
warmup: 8%|▊ | 75/1000 [00:32<09:58, 1.55it/s, 1023 steps of size 5.09e-03. acc. prob=0.75]
warmup: 8%|▊ | 76/1000 [00:33<10:20, 1.49it/s, 1023 steps of size 6.81e-03. acc. prob=0.75]
warmup: 8%|▊ | 77/1000 [00:33<08:53, 1.73it/s, 511 steps of size 2.39e-03. acc. prob=0.75]
warmup: 8%|▊ | 78/1000 [00:34<09:33, 1.61it/s, 1023 steps of size 3.12e-03. acc. prob=0.75]
warmup: 8%|▊ | 79/1000 [00:34<09:59, 1.54it/s, 1023 steps of size 4.25e-03. acc. prob=0.75]
warmup: 8%|▊ | 80/1000 [00:35<08:44, 1.75it/s, 511 steps of size 6.56e-03. acc. prob=0.75]
warmup: 8%|▊ | 81/1000 [00:35<07:44, 1.98it/s, 511 steps of size 3.97e-03. acc. prob=0.75]
warmup: 8%|▊ | 82/1000 [00:36<08:44, 1.75it/s, 1023 steps of size 6.12e-03. acc. prob=0.75]
warmup: 8%|▊ | 83/1000 [00:36<07:41, 1.99it/s, 511 steps of size 2.30e-03. acc. prob=0.75]
warmup: 8%|▊ | 84/1000 [00:37<08:13, 1.85it/s, 1023 steps of size 3.23e-03. acc. prob=0.75]
warmup: 8%|▊ | 85/1000 [00:38<08:32, 1.79it/s, 1023 steps of size 4.90e-03. acc. prob=0.75]
warmup: 9%|▊ | 86/1000 [00:38<08:48, 1.73it/s, 1023 steps of size 3.98e-03. acc. prob=0.75]
warmup: 9%|▊ | 87/1000 [00:39<08:42, 1.75it/s, 1023 steps of size 5.55e-03. acc. prob=0.76]
warmup: 9%|▉ | 88/1000 [00:39<07:22, 2.06it/s, 511 steps of size 8.06e-03. acc. prob=0.76]
warmup: 9%|▉ | 89/1000 [00:39<06:25, 2.36it/s, 511 steps of size 2.39e-03. acc. prob=0.75]
warmup: 9%|▉ | 90/1000 [00:40<07:03, 2.15it/s, 1023 steps of size 3.62e-03. acc. prob=0.75]
warmup: 9%|▉ | 91/1000 [00:40<07:30, 2.02it/s, 1023 steps of size 4.55e-03. acc. prob=0.76]
warmup: 9%|▉ | 92/1000 [00:41<07:50, 1.93it/s, 1023 steps of size 4.77e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:41<06:47, 2.23it/s, 511 steps of size 6.34e-03. acc. prob=0.76]
warmup: 9%|▉ | 94/1000 [00:42<06:03, 2.49it/s, 511 steps of size 4.57e-03. acc. prob=0.76]
warmup: 10%|▉ | 95/1000 [00:42<06:48, 2.21it/s, 1023 steps of size 6.83e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:42<06:04, 2.48it/s, 511 steps of size 8.47e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:43<05:32, 2.71it/s, 511 steps of size 8.95e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:43<04:12, 3.57it/s, 511 steps of size 5.79e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:43<04:13, 3.56it/s, 511 steps of size 6.68e-03. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:44<04:14, 3.53it/s, 511 steps of size 9.60e-02. acc. prob=0.76]
warmup: 10%|█ | 105/1000 [00:44<01:57, 7.65it/s, 127 steps of size 5.86e-02. acc. prob=0.76]
warmup: 11%|█ | 109/1000 [00:44<01:26, 10.28it/s, 255 steps of size 4.66e-02. acc. prob=0.76]
warmup: 11%|█ | 111/1000 [00:44<01:16, 11.56it/s, 63 steps of size 1.45e-01. acc. prob=0.77]
warmup: 12%|█▏ | 115/1000 [00:44<00:54, 16.17it/s, 63 steps of size 9.40e-02. acc. prob=0.77]
warmup: 12%|█▏ | 119/1000 [00:45<01:02, 14.13it/s, 511 steps of size 3.61e-02. acc. prob=0.76]
warmup: 12%|█▏ | 121/1000 [00:45<01:02, 14.11it/s, 127 steps of size 7.93e-02. acc. prob=0.77]
warmup: 12%|█▏ | 124/1000 [00:45<01:13, 11.94it/s, 511 steps of size 2.40e-02. acc. prob=0.76]
warmup: 12%|█▎ | 125/1000 [00:45<01:20, 10.92it/s, 255 steps of size 4.25e-02. acc. prob=0.77]
warmup: 13%|█▎ | 127/1000 [00:45<01:10, 12.32it/s, 63 steps of size 7.98e-02. acc. prob=0.77]
warmup: 13%|█▎ | 130/1000 [00:45<00:56, 15.50it/s, 63 steps of size 1.06e-01. acc. prob=0.77]
warmup: 13%|█▎ | 133/1000 [00:46<00:56, 15.44it/s, 255 steps of size 3.37e-02. acc. prob=0.77]
warmup: 14%|█▎ | 135/1000 [00:46<00:57, 15.00it/s, 127 steps of size 8.20e-02. acc. prob=0.77]
warmup: 14%|█▍ | 140/1000 [00:46<00:48, 17.77it/s, 255 steps of size 4.42e-02. acc. prob=0.77]
warmup: 14%|█▍ | 142/1000 [00:46<00:48, 17.85it/s, 63 steps of size 9.59e-02. acc. prob=0.77]
warmup: 15%|█▍ | 146/1000 [00:46<00:38, 21.93it/s, 31 steps of size 1.72e-01. acc. prob=0.78]
warmup: 15%|█▍ | 148/1000 [00:46<00:45, 18.83it/s, 255 steps of size 4.35e-02. acc. prob=0.77]
warmup: 15%|█▌ | 150/1000 [00:46<00:45, 18.79it/s, 63 steps of size 1.05e-01. acc. prob=0.77]
warmup: 15%|█▌ | 154/1000 [00:47<00:43, 19.39it/s, 255 steps of size 3.20e-02. acc. prob=0.77]
warmup: 16%|█▌ | 155/1000 [00:47<00:53, 15.93it/s, 255 steps of size 4.93e-02. acc. prob=0.77]
warmup: 16%|█▌ | 157/1000 [00:47<00:50, 16.62it/s, 63 steps of size 1.04e-01. acc. prob=0.77]
warmup: 16%|█▌ | 160/1000 [00:47<00:42, 19.56it/s, 63 steps of size 1.13e-01. acc. prob=0.77]
warmup: 16%|█▋ | 163/1000 [00:47<00:44, 18.68it/s, 255 steps of size 3.77e-02. acc. prob=0.77]
warmup: 16%|█▋ | 164/1000 [00:47<00:54, 15.21it/s, 255 steps of size 6.51e-02. acc. prob=0.77]
warmup: 17%|█▋ | 168/1000 [00:48<00:52, 15.77it/s, 255 steps of size 2.68e-02. acc. prob=0.77]
warmup: 17%|█▋ | 169/1000 [00:48<01:01, 13.47it/s, 255 steps of size 5.03e-02. acc. prob=0.77]
warmup: 17%|█▋ | 171/1000 [00:48<00:56, 14.64it/s, 63 steps of size 1.49e-01. acc. prob=0.77]
warmup: 17%|█▋ | 174/1000 [00:48<00:52, 15.83it/s, 255 steps of size 4.54e-02. acc. prob=0.77]
warmup: 18%|█▊ | 176/1000 [00:48<00:49, 16.53it/s, 63 steps of size 1.30e-01. acc. prob=0.77]
warmup: 18%|█▊ | 180/1000 [00:48<00:37, 21.62it/s, 31 steps of size 2.08e-01. acc. prob=0.78]
warmup: 18%|█▊ | 183/1000 [00:48<00:38, 21.17it/s, 127 steps of size 8.57e-02. acc. prob=0.77]
warmup: 19%|█▊ | 186/1000 [00:48<00:37, 22.00it/s, 127 steps of size 5.54e-02. acc. prob=0.77]
warmup: 19%|█▉ | 190/1000 [00:49<00:32, 24.58it/s, 63 steps of size 1.38e-01. acc. prob=0.78]
warmup: 19%|█▉ | 192/1000 [00:49<00:39, 20.46it/s, 255 steps of size 3.20e-02. acc. prob=0.77]
warmup: 19%|█▉ | 194/1000 [00:49<00:43, 18.59it/s, 127 steps of size 7.90e-02. acc. prob=0.77]
warmup: 20%|█▉ | 197/1000 [00:49<00:38, 20.98it/s, 63 steps of size 1.64e-01. acc. prob=0.78]
warmup: 20%|██ | 200/1000 [00:49<00:39, 20.30it/s, 127 steps of size 9.57e-02. acc. prob=0.78]
warmup: 20%|██ | 204/1000 [00:49<00:32, 24.63it/s, 63 steps of size 8.38e-02. acc. prob=0.78]
warmup: 21%|██ | 208/1000 [00:49<00:28, 28.03it/s, 31 steps of size 1.02e-01. acc. prob=0.78]
warmup: 21%|██ | 212/1000 [00:49<00:27, 29.09it/s, 127 steps of size 6.16e-02. acc. prob=0.78]
warmup: 21%|██▏ | 214/1000 [00:50<00:29, 26.33it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 22%|██▏ | 217/1000 [00:50<00:30, 25.71it/s, 127 steps of size 8.35e-02. acc. prob=0.78]
warmup: 22%|██▏ | 221/1000 [00:50<00:28, 27.63it/s, 63 steps of size 1.46e-01. acc. prob=0.78]
warmup: 22%|██▏ | 224/1000 [00:50<00:29, 26.47it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 23%|██▎ | 228/1000 [00:50<00:27, 28.29it/s, 63 steps of size 9.55e-02. acc. prob=0.78]
warmup: 23%|██▎ | 232/1000 [00:50<00:26, 29.21it/s, 127 steps of size 6.20e-02. acc. prob=0.78]
warmup: 24%|██▎ | 236/1000 [00:50<00:24, 31.47it/s, 31 steps of size 9.98e-02. acc. prob=0.78]
warmup: 24%|██▍ | 240/1000 [00:50<00:22, 33.23it/s, 63 steps of size 1.28e-01. acc. prob=0.78]
warmup: 24%|██▍ | 244/1000 [00:50<00:21, 34.44it/s, 31 steps of size 1.17e-01. acc. prob=0.78]
warmup: 25%|██▍ | 248/1000 [00:51<00:21, 35.34it/s, 63 steps of size 8.55e-02. acc. prob=0.78]
warmup: 25%|██▌ | 254/1000 [00:51<00:25, 28.88it/s, 255 steps of size 3.15e-02. acc. prob=0.78]
warmup: 26%|██▌ | 256/1000 [00:51<00:30, 24.55it/s, 127 steps of size 7.45e-02. acc. prob=0.78]
warmup: 26%|██▌ | 260/1000 [00:51<00:32, 22.91it/s, 255 steps of size 3.41e-02. acc. prob=0.78]
warmup: 26%|██▌ | 262/1000 [00:51<00:33, 21.91it/s, 63 steps of size 9.05e-02. acc. prob=0.78]
warmup: 27%|██▋ | 266/1000 [00:52<00:35, 20.47it/s, 255 steps of size 2.96e-02. acc. prob=0.78]
warmup: 27%|██▋ | 268/1000 [00:52<00:36, 20.01it/s, 63 steps of size 6.67e-02. acc. prob=0.78]
warmup: 27%|██▋ | 272/1000 [00:52<00:32, 22.71it/s, 127 steps of size 6.92e-02. acc. prob=0.78]
warmup: 28%|██▊ | 276/1000 [00:52<00:35, 20.30it/s, 255 steps of size 3.31e-02. acc. prob=0.78]
warmup: 28%|██▊ | 278/1000 [00:52<00:36, 19.80it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 28%|██▊ | 281/1000 [00:52<00:37, 19.05it/s, 255 steps of size 3.40e-02. acc. prob=0.78]
warmup: 28%|██▊ | 283/1000 [00:52<00:38, 18.84it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 29%|██▊ | 287/1000 [00:53<00:32, 22.26it/s, 63 steps of size 1.75e-01. acc. prob=0.78]
warmup: 29%|██▉ | 290/1000 [00:53<00:30, 23.34it/s, 63 steps of size 1.01e-01. acc. prob=0.78]
warmup: 29%|██▉ | 294/1000 [00:53<00:29, 23.60it/s, 127 steps of size 8.39e-02. acc. prob=0.78]
warmup: 30%|██▉ | 297/1000 [00:53<00:33, 21.06it/s, 255 steps of size 2.82e-02. acc. prob=0.78]
warmup: 30%|██▉ | 299/1000 [00:53<00:37, 18.80it/s, 127 steps of size 6.35e-02. acc. prob=0.78]
warmup: 30%|███ | 303/1000 [00:53<00:30, 22.86it/s, 31 steps of size 1.35e-01. acc. prob=0.78]
warmup: 31%|███ | 306/1000 [00:53<00:32, 21.29it/s, 127 steps of size 6.88e-02. acc. prob=0.78]
warmup: 31%|███ | 310/1000 [00:54<00:27, 24.80it/s, 31 steps of size 1.22e-01. acc. prob=0.78]
warmup: 31%|███▏ | 313/1000 [00:54<00:28, 24.34it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 32%|███▏ | 317/1000 [00:54<00:28, 24.02it/s, 127 steps of size 8.88e-02. acc. prob=0.78]
warmup: 32%|███▏ | 321/1000 [00:54<00:26, 26.04it/s, 63 steps of size 3.83e-02. acc. prob=0.78]
warmup: 32%|███▏ | 323/1000 [00:54<00:28, 24.03it/s, 63 steps of size 7.96e-02. acc. prob=0.78]
warmup: 32%|███▎ | 325/1000 [00:54<00:29, 22.51it/s, 127 steps of size 8.73e-02. acc. prob=0.78]
warmup: 33%|███▎ | 329/1000 [00:54<00:25, 26.44it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 33%|███▎ | 332/1000 [00:54<00:24, 27.05it/s, 127 steps of size 4.89e-02. acc. prob=0.78]
warmup: 33%|███▎ | 334/1000 [00:55<00:27, 24.46it/s, 63 steps of size 9.76e-02. acc. prob=0.78]
warmup: 34%|███▍ | 338/1000 [00:55<00:26, 25.40it/s, 127 steps of size 7.29e-02. acc. prob=0.78]
warmup: 34%|███▍ | 341/1000 [00:55<00:25, 25.92it/s, 63 steps of size 1.32e-01. acc. prob=0.78]
warmup: 34%|███▍ | 345/1000 [00:55<00:22, 29.55it/s, 63 steps of size 3.23e-02. acc. prob=0.78]
warmup: 35%|███▍ | 347/1000 [00:55<00:27, 23.91it/s, 127 steps of size 5.47e-02. acc. prob=0.78]
warmup: 35%|███▍ | 349/1000 [00:55<00:29, 22.38it/s, 63 steps of size 9.61e-02. acc. prob=0.78]
warmup: 35%|███▌ | 352/1000 [00:55<00:27, 23.73it/s, 63 steps of size 7.54e-02. acc. prob=0.78]
warmup: 36%|███▌ | 356/1000 [00:55<00:26, 24.00it/s, 127 steps of size 5.53e-02. acc. prob=0.78]
warmup: 36%|███▌ | 359/1000 [00:56<00:25, 24.87it/s, 63 steps of size 1.13e-01. acc. prob=0.78]
warmup: 36%|███▋ | 363/1000 [00:56<00:22, 28.08it/s, 31 steps of size 1.01e-01. acc. prob=0.78]
warmup: 37%|███▋ | 367/1000 [00:56<00:21, 29.06it/s, 63 steps of size 1.14e-01. acc. prob=0.78]
warmup: 37%|███▋ | 371/1000 [00:56<00:21, 29.61it/s, 63 steps of size 6.36e-02. acc. prob=0.78]
warmup: 37%|███▋ | 374/1000 [00:56<00:21, 28.97it/s, 63 steps of size 9.97e-02. acc. prob=0.78]
warmup: 38%|███▊ | 378/1000 [00:56<00:20, 29.75it/s, 63 steps of size 9.92e-02. acc. prob=0.78]
warmup: 38%|███▊ | 382/1000 [00:56<00:19, 32.24it/s, 63 steps of size 6.67e-02. acc. prob=0.78]
warmup: 39%|███▊ | 386/1000 [00:56<00:18, 33.40it/s, 31 steps of size 1.47e-01. acc. prob=0.78]
warmup: 39%|███▉ | 391/1000 [00:56<00:16, 36.81it/s, 31 steps of size 9.52e-02. acc. prob=0.78]
warmup: 40%|███▉ | 395/1000 [00:57<00:17, 34.67it/s, 127 steps of size 6.51e-02. acc. prob=0.78]
warmup: 40%|███▉ | 398/1000 [00:57<00:18, 32.61it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 40%|████ | 403/1000 [00:57<00:17, 34.36it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 41%|████ | 408/1000 [00:57<00:16, 36.59it/s, 63 steps of size 6.79e-02. acc. prob=0.78]
warmup: 41%|████ | 411/1000 [00:57<00:17, 33.92it/s, 63 steps of size 7.01e-02. acc. prob=0.78]
warmup: 42%|████▏ | 415/1000 [00:57<00:17, 32.97it/s, 63 steps of size 1.22e-01. acc. prob=0.78]
warmup: 42%|████▏ | 418/1000 [00:57<00:18, 31.18it/s, 127 steps of size 6.47e-02. acc. prob=0.78]
warmup: 42%|████▏ | 422/1000 [00:57<00:20, 28.56it/s, 127 steps of size 6.33e-02. acc. prob=0.78]
warmup: 42%|████▎ | 425/1000 [00:58<00:20, 28.21it/s, 63 steps of size 9.72e-02. acc. prob=0.78]
warmup: 43%|████▎ | 430/1000 [00:58<00:18, 30.96it/s, 63 steps of size 9.94e-02. acc. prob=0.78]
warmup: 43%|████▎ | 434/1000 [00:58<00:17, 32.40it/s, 31 steps of size 5.73e-02. acc. prob=0.78]
warmup: 44%|████▎ | 437/1000 [00:58<00:19, 28.28it/s, 127 steps of size 8.14e-02. acc. prob=0.78]
warmup: 44%|████▍ | 441/1000 [00:58<00:18, 30.50it/s, 31 steps of size 1.41e-01. acc. prob=0.79]
warmup: 44%|████▍ | 444/1000 [00:58<00:19, 28.34it/s, 63 steps of size 8.65e-02. acc. prob=0.78]
warmup: 45%|████▍ | 448/1000 [00:58<00:18, 29.22it/s, 63 steps of size 9.13e-02. acc. prob=0.78]
warmup: 45%|████▌ | 451/1000 [00:58<00:19, 28.76it/s, 63 steps of size 1.03e+00. acc. prob=0.78]
warmup: 46%|████▌ | 456/1000 [00:59<00:16, 33.35it/s, 31 steps of size 9.91e-02. acc. prob=0.78]
warmup: 46%|████▌ | 459/1000 [00:59<00:17, 31.62it/s, 63 steps of size 1.61e-01. acc. prob=0.78]
warmup: 46%|████▌ | 461/1000 [00:59<00:21, 24.75it/s, 255 steps of size 2.92e-02. acc. prob=0.78]
warmup: 46%|████▋ | 463/1000 [00:59<00:23, 22.97it/s, 63 steps of size 1.03e-01. acc. prob=0.78]
warmup: 47%|████▋ | 467/1000 [00:59<00:21, 24.48it/s, 127 steps of size 8.14e-02. acc. prob=0.78]
warmup: 47%|████▋ | 471/1000 [00:59<00:23, 22.58it/s, 255 steps of size 4.09e-02. acc. prob=0.78]
warmup: 47%|████▋ | 473/1000 [00:59<00:24, 21.59it/s, 63 steps of size 1.27e-01. acc. prob=0.78]
warmup: 48%|████▊ | 476/1000 [00:59<00:23, 22.11it/s, 63 steps of size 7.93e-02. acc. prob=0.78]
warmup: 48%|████▊ | 479/1000 [01:00<00:23, 22.47it/s, 127 steps of size 4.14e-02. acc. prob=0.78]
warmup: 48%|████▊ | 481/1000 [01:00<00:24, 21.34it/s, 63 steps of size 1.26e-01. acc. prob=0.78]
warmup: 48%|████▊ | 484/1000 [01:00<00:23, 21.99it/s, 127 steps of size 6.82e-02. acc. prob=0.78]
warmup: 49%|████▉ | 488/1000 [01:00<00:25, 20.06it/s, 255 steps of size 3.87e-02. acc. prob=0.78]
warmup: 49%|████▉ | 490/1000 [01:00<00:26, 19.60it/s, 63 steps of size 7.14e-02. acc. prob=0.78]
warmup: 49%|████▉ | 494/1000 [01:00<00:26, 19.10it/s, 255 steps of size 4.05e-02. acc. prob=0.78]
warmup: 50%|████▉ | 496/1000 [01:01<00:26, 18.89it/s, 63 steps of size 8.17e-02. acc. prob=0.78]
warmup: 50%|████▉ | 499/1000 [01:01<00:23, 20.90it/s, 63 steps of size 1.29e-01. acc. prob=0.78]
sample: 50%|█████ | 503/1000 [01:01<00:20, 23.82it/s, 63 steps of size 6.61e-02. acc. prob=0.98]
sample: 51%|█████ | 506/1000 [01:01<00:20, 24.65it/s, 63 steps of size 6.61e-02. acc. prob=0.92]
sample: 51%|█████ | 509/1000 [01:01<00:19, 25.39it/s, 63 steps of size 6.61e-02. acc. prob=0.93]
sample: 51%|█████ | 512/1000 [01:01<00:18, 25.89it/s, 63 steps of size 6.61e-02. acc. prob=0.93]
sample: 52%|█████▏ | 515/1000 [01:01<00:18, 26.33it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 52%|█████▏ | 518/1000 [01:01<00:18, 26.61it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 52%|█████▏ | 521/1000 [01:01<00:17, 26.72it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 52%|█████▏ | 524/1000 [01:02<00:17, 26.85it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 53%|█████▎ | 527/1000 [01:02<00:17, 27.03it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 53%|█████▎ | 530/1000 [01:02<00:17, 27.10it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 53%|█████▎ | 533/1000 [01:02<00:17, 27.11it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 54%|█████▎ | 536/1000 [01:02<00:17, 27.14it/s, 63 steps of size 6.61e-02. acc. prob=0.96]
sample: 54%|█████▍ | 539/1000 [01:02<00:16, 27.18it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 54%|█████▍ | 542/1000 [01:02<00:16, 27.19it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 54%|█████▍ | 544/1000 [01:02<00:18, 24.45it/s, 127 steps of size 6.61e-02. acc. prob=0.95]
sample: 55%|█████▍ | 547/1000 [01:02<00:17, 25.33it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 55%|█████▌ | 550/1000 [01:03<00:17, 25.98it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 55%|█████▌ | 553/1000 [01:03<00:16, 26.44it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 56%|█████▌ | 556/1000 [01:03<00:16, 26.68it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 56%|█████▌ | 559/1000 [01:03<00:16, 26.87it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 56%|█████▌ | 562/1000 [01:03<00:16, 26.96it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 56%|█████▋ | 565/1000 [01:03<00:16, 27.05it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 57%|█████▋ | 568/1000 [01:03<00:15, 27.19it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 57%|█████▋ | 571/1000 [01:03<00:15, 27.29it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 57%|█████▋ | 574/1000 [01:03<00:15, 27.27it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 58%|█████▊ | 577/1000 [01:04<00:15, 27.32it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 58%|█████▊ | 580/1000 [01:04<00:15, 27.34it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 58%|█████▊ | 583/1000 [01:04<00:15, 27.40it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 59%|█████▊ | 586/1000 [01:04<00:15, 27.32it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 59%|█████▉ | 589/1000 [01:04<00:15, 27.30it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 59%|█████▉ | 592/1000 [01:04<00:14, 27.34it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 60%|█████▉ | 595/1000 [01:04<00:14, 27.38it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 60%|█████▉ | 598/1000 [01:04<00:14, 27.40it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 60%|██████ | 601/1000 [01:04<00:14, 27.35it/s, 63 steps of size 6.61e-02. acc. prob=0.95]
sample: 60%|██████ | 604/1000 [01:05<00:14, 27.30it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 61%|██████ | 607/1000 [01:05<00:14, 27.29it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 61%|██████ | 610/1000 [01:05<00:14, 27.34it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 61%|██████▏ | 613/1000 [01:05<00:14, 27.32it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 62%|██████▏ | 616/1000 [01:05<00:14, 27.29it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 62%|██████▏ | 619/1000 [01:05<00:13, 27.25it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 62%|██████▏ | 622/1000 [01:05<00:13, 27.29it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 62%|██████▎ | 625/1000 [01:05<00:13, 27.32it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 63%|██████▎ | 628/1000 [01:05<00:13, 27.33it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 63%|██████▎ | 631/1000 [01:05<00:13, 27.25it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 63%|██████▎ | 634/1000 [01:06<00:13, 27.29it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 64%|██████▎ | 637/1000 [01:06<00:13, 27.30it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 64%|██████▍ | 640/1000 [01:06<00:13, 27.29it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 64%|██████▍ | 643/1000 [01:06<00:13, 27.28it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 65%|██████▍ | 646/1000 [01:06<00:12, 27.28it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 65%|██████▍ | 649/1000 [01:06<00:12, 27.18it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 65%|██████▌ | 652/1000 [01:06<00:12, 27.23it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 66%|██████▌ | 655/1000 [01:06<00:12, 27.23it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 66%|██████▌ | 658/1000 [01:06<00:12, 27.31it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 66%|██████▌ | 661/1000 [01:07<00:12, 27.31it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 66%|██████▋ | 664/1000 [01:07<00:12, 27.32it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 67%|██████▋ | 667/1000 [01:07<00:12, 27.31it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 67%|██████▋ | 670/1000 [01:07<00:12, 27.30it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 67%|██████▋ | 673/1000 [01:07<00:11, 27.29it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 68%|██████▊ | 676/1000 [01:07<00:11, 27.36it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 68%|██████▊ | 678/1000 [01:07<00:13, 24.59it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 68%|██████▊ | 681/1000 [01:07<00:12, 25.46it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 68%|██████▊ | 684/1000 [01:07<00:12, 25.90it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 69%|██████▊ | 687/1000 [01:08<00:11, 26.24it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 69%|██████▉ | 690/1000 [01:08<00:11, 26.46it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 69%|██████▉ | 693/1000 [01:08<00:11, 26.68it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 70%|██████▉ | 696/1000 [01:08<00:11, 26.87it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 70%|██████▉ | 699/1000 [01:08<00:11, 26.99it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 70%|███████ | 702/1000 [01:08<00:11, 27.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 70%|███████ | 705/1000 [01:08<00:10, 27.17it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 71%|███████ | 708/1000 [01:08<00:10, 27.18it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 71%|███████ | 711/1000 [01:08<00:10, 27.20it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 71%|███████▏ | 714/1000 [01:09<00:10, 27.16it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 72%|███████▏ | 717/1000 [01:09<00:10, 27.21it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 72%|███████▏ | 720/1000 [01:09<00:10, 27.13it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 72%|███████▏ | 723/1000 [01:09<00:10, 27.14it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 72%|███████▎ | 725/1000 [01:09<00:11, 24.49it/s, 127 steps of size 6.61e-02. acc. prob=0.94]
sample: 73%|███████▎ | 728/1000 [01:09<00:10, 25.53it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 73%|███████▎ | 731/1000 [01:09<00:10, 26.21it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 73%|███████▎ | 734/1000 [01:09<00:11, 22.24it/s, 191 steps of size 6.61e-02. acc. prob=0.94]
sample: 74%|███████▎ | 736/1000 [01:10<00:12, 21.32it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 74%|███████▍ | 738/1000 [01:10<00:12, 20.65it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 74%|███████▍ | 741/1000 [01:10<00:11, 22.63it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 74%|███████▍ | 744/1000 [01:10<00:10, 24.15it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 75%|███████▍ | 747/1000 [01:10<00:10, 25.20it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 75%|███████▌ | 750/1000 [01:10<00:09, 26.05it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 75%|███████▌ | 753/1000 [01:10<00:09, 26.61it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 76%|███████▌ | 756/1000 [01:10<00:09, 26.98it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 76%|███████▌ | 759/1000 [01:10<00:08, 27.23it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 76%|███████▌ | 762/1000 [01:10<00:08, 27.42it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 76%|███████▋ | 765/1000 [01:11<00:08, 27.57it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 77%|███████▋ | 768/1000 [01:11<00:08, 27.83it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 77%|███████▋ | 771/1000 [01:11<00:08, 27.82it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 77%|███████▋ | 774/1000 [01:11<00:08, 25.30it/s, 127 steps of size 6.61e-02. acc. prob=0.94]
sample: 78%|███████▊ | 777/1000 [01:11<00:08, 26.10it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 78%|███████▊ | 780/1000 [01:11<00:08, 26.74it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 78%|███████▊ | 783/1000 [01:11<00:08, 27.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 79%|███████▊ | 786/1000 [01:11<00:07, 27.36it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 79%|███████▉ | 789/1000 [01:11<00:07, 27.57it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 79%|███████▉ | 792/1000 [01:12<00:07, 27.70it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 80%|███████▉ | 795/1000 [01:12<00:07, 27.79it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 80%|███████▉ | 798/1000 [01:12<00:07, 27.86it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 80%|████████ | 801/1000 [01:12<00:07, 27.89it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 80%|████████ | 804/1000 [01:12<00:07, 27.92it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 81%|████████ | 807/1000 [01:12<00:06, 27.94it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 81%|████████ | 810/1000 [01:12<00:06, 27.90it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 81%|████████▏ | 813/1000 [01:12<00:06, 27.98it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 82%|████████▏ | 816/1000 [01:12<00:06, 28.07it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 82%|████████▏ | 819/1000 [01:13<00:06, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 82%|████████▏ | 822/1000 [01:13<00:06, 28.18it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 82%|████████▎ | 825/1000 [01:13<00:06, 28.24it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 83%|████████▎ | 828/1000 [01:13<00:06, 28.19it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 83%|████████▎ | 831/1000 [01:13<00:05, 28.27it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 83%|████████▎ | 834/1000 [01:13<00:05, 28.23it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 84%|████████▎ | 837/1000 [01:13<00:05, 28.12it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 84%|████████▍ | 840/1000 [01:13<00:05, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 84%|████████▍ | 843/1000 [01:13<00:05, 28.05it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 85%|████████▍ | 846/1000 [01:14<00:05, 27.93it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 85%|████████▍ | 849/1000 [01:14<00:05, 28.02it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 85%|████████▌ | 852/1000 [01:14<00:05, 28.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 86%|████████▌ | 855/1000 [01:14<00:05, 28.09it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 86%|████████▌ | 858/1000 [01:14<00:05, 28.15it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 86%|████████▌ | 861/1000 [01:14<00:04, 28.14it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 86%|████████▋ | 864/1000 [01:14<00:04, 28.13it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 87%|████████▋ | 867/1000 [01:14<00:04, 28.10it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 87%|████████▋ | 870/1000 [01:14<00:04, 28.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 87%|████████▋ | 873/1000 [01:14<00:04, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 88%|████████▊ | 876/1000 [01:15<00:04, 28.06it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 88%|████████▊ | 879/1000 [01:15<00:04, 28.09it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 88%|████████▊ | 882/1000 [01:15<00:04, 28.05it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 88%|████████▊ | 885/1000 [01:15<00:04, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 89%|████████▉ | 888/1000 [01:15<00:03, 28.18it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 89%|████████▉ | 891/1000 [01:15<00:03, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 89%|████████▉ | 894/1000 [01:15<00:03, 28.12it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 90%|████████▉ | 897/1000 [01:15<00:03, 28.06it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 90%|█████████ | 900/1000 [01:15<00:03, 28.00it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 90%|█████████ | 903/1000 [01:16<00:03, 28.07it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 91%|█████████ | 906/1000 [01:16<00:03, 28.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 91%|█████████ | 909/1000 [01:16<00:03, 28.13it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 91%|█████████ | 912/1000 [01:16<00:03, 28.14it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 92%|█████████▏| 915/1000 [01:16<00:03, 28.07it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 92%|█████████▏| 918/1000 [01:16<00:02, 28.16it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 92%|█████████▏| 921/1000 [01:16<00:02, 28.13it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 92%|█████████▏| 924/1000 [01:16<00:02, 28.14it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 93%|█████████▎| 927/1000 [01:16<00:02, 28.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 93%|█████████▎| 930/1000 [01:17<00:02, 28.07it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 93%|█████████▎| 933/1000 [01:17<00:02, 28.07it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 94%|█████████▎| 936/1000 [01:17<00:02, 28.07it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 94%|█████████▍| 939/1000 [01:17<00:02, 28.01it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 94%|█████████▍| 942/1000 [01:17<00:02, 28.06it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 94%|█████████▍| 945/1000 [01:17<00:01, 28.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 95%|█████████▍| 948/1000 [01:17<00:01, 27.96it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 95%|█████████▌| 951/1000 [01:17<00:01, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 95%|█████████▌| 954/1000 [01:17<00:01, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 96%|█████████▌| 957/1000 [01:17<00:01, 27.89it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 96%|█████████▌| 960/1000 [01:18<00:01, 27.87it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 96%|█████████▋| 963/1000 [01:18<00:01, 27.79it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 97%|█████████▋| 966/1000 [01:18<00:01, 27.89it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 97%|█████████▋| 969/1000 [01:18<00:01, 27.97it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 97%|█████████▋| 972/1000 [01:18<00:01, 27.97it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 98%|█████████▊| 975/1000 [01:18<00:00, 28.03it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 98%|█████████▊| 978/1000 [01:18<00:00, 28.04it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 98%|█████████▊| 981/1000 [01:18<00:00, 28.07it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 98%|█████████▊| 984/1000 [01:18<00:00, 28.08it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 99%|█████████▊| 987/1000 [01:19<00:00, 28.06it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 99%|█████████▉| 990/1000 [01:19<00:00, 28.12it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 99%|█████████▉| 993/1000 [01:19<00:00, 28.09it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 100%|█████████▉| 996/1000 [01:19<00:00, 28.11it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 100%|█████████▉| 999/1000 [01:19<00:00, 28.20it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
sample: 100%|██████████| 1000/1000 [01:19<00:00, 12.58it/s, 63 steps of size 6.61e-02. acc. prob=0.94]
0%| | 0/1000 [00:00<?, ?it/s]
warmup: 5%|▌ | 50/1000 [00:12<03:58, 3.98it/s, 1023 steps of size 2.94e-03. acc. prob=0.73]
warmup: 5%|▌ | 50/1000 [00:27<03:58, 3.98it/s, 511 steps of size 4.22e-03. acc. prob=0.75]
warmup: 8%|▊ | 84/1000 [00:27<05:10, 2.95it/s, 1023 steps of size 4.01e-03. acc. prob=0.75]
warmup: 8%|▊ | 85/1000 [00:27<05:14, 2.91it/s, 1023 steps of size 6.00e-03. acc. prob=0.76]
warmup: 9%|▊ | 86/1000 [00:28<05:12, 2.92it/s, 511 steps of size 8.73e-03. acc. prob=0.76]
warmup: 9%|▊ | 87/1000 [00:28<05:04, 3.00it/s, 217 steps of size 1.96e-03. acc. prob=0.75]
warmup: 9%|▉ | 88/1000 [00:28<05:15, 2.89it/s, 1023 steps of size 2.96e-03. acc. prob=0.75]
warmup: 9%|▉ | 89/1000 [00:29<05:29, 2.76it/s, 1023 steps of size 4.49e-03. acc. prob=0.75]
warmup: 9%|▉ | 90/1000 [00:29<05:45, 2.63it/s, 1023 steps of size 6.53e-03. acc. prob=0.76]
warmup: 9%|▉ | 91/1000 [00:30<05:35, 2.71it/s, 511 steps of size 9.19e-03. acc. prob=0.76]
warmup: 9%|▉ | 93/1000 [00:30<05:17, 2.85it/s, 1023 steps of size 3.04e-03. acc. prob=0.75]
warmup: 9%|▉ | 94/1000 [00:31<05:47, 2.61it/s, 1023 steps of size 4.52e-03. acc. prob=0.76]
warmup: 10%|▉ | 95/1000 [00:31<06:16, 2.40it/s, 1023 steps of size 5.48e-03. acc. prob=0.76]
warmup: 10%|▉ | 96/1000 [00:32<05:54, 2.55it/s, 511 steps of size 6.40e-03. acc. prob=0.76]
warmup: 10%|▉ | 97/1000 [00:32<05:35, 2.69it/s, 511 steps of size 4.96e-03. acc. prob=0.76]
warmup: 10%|▉ | 98/1000 [00:33<06:23, 2.35it/s, 1023 steps of size 5.16e-03. acc. prob=0.76]
warmup: 10%|▉ | 99/1000 [00:33<05:53, 2.55it/s, 511 steps of size 7.31e-03. acc. prob=0.76]
warmup: 10%|█ | 100/1000 [00:33<05:30, 2.72it/s, 511 steps of size 1.05e-02. acc. prob=0.76]
warmup: 10%|█ | 101/1000 [00:33<05:12, 2.87it/s, 511 steps of size 1.51e-01. acc. prob=0.76]
warmup: 10%|█ | 103/1000 [00:34<03:24, 4.39it/s, 255 steps of size 5.30e-02. acc. prob=0.76]
warmup: 10%|█ | 105/1000 [00:34<02:24, 6.19it/s, 63 steps of size 1.23e-01. acc. prob=0.76]
warmup: 11%|█ | 107/1000 [00:34<01:50, 8.08it/s, 127 steps of size 5.23e-02. acc. prob=0.76]
warmup: 11%|█ | 109/1000 [00:34<01:29, 9.91it/s, 127 steps of size 1.11e-01. acc. prob=0.77]
warmup: 11%|█ | 111/1000 [00:34<01:24, 10.47it/s, 255 steps of size 4.88e-02. acc. prob=0.76]
warmup: 11%|█▏ | 113/1000 [00:34<01:18, 11.25it/s, 127 steps of size 8.18e-02. acc. prob=0.77]
warmup: 12%|█▏ | 116/1000 [00:34<01:00, 14.63it/s, 127 steps of size 6.54e-02. acc. prob=0.77]
warmup: 12%|█▏ | 119/1000 [00:34<00:50, 17.54it/s, 63 steps of size 1.23e-02. acc. prob=0.76]
warmup: 12%|█▏ | 120/1000 [00:35<01:21, 10.79it/s, 511 steps of size 2.28e-02. acc. prob=0.76]
warmup: 12%|█▏ | 121/1000 [00:35<01:30, 9.75it/s, 255 steps of size 3.69e-02. acc. prob=0.77]
warmup: 12%|█▏ | 123/1000 [00:35<01:21, 10.75it/s, 127 steps of size 1.03e-01. acc. prob=0.77]
warmup: 12%|█▎ | 125/1000 [00:35<01:10, 12.35it/s, 127 steps of size 9.06e-02. acc. prob=0.77]
warmup: 13%|█▎ | 128/1000 [00:35<01:08, 12.74it/s, 255 steps of size 2.49e-02. acc. prob=0.77]
warmup: 13%|█▎ | 129/1000 [00:36<01:18, 11.13it/s, 255 steps of size 4.30e-02. acc. prob=0.77]
warmup: 13%|█▎ | 131/1000 [00:36<01:13, 11.81it/s, 127 steps of size 9.75e-02. acc. prob=0.77]
warmup: 13%|█▎ | 134/1000 [00:36<01:00, 14.25it/s, 127 steps of size 1.02e-01. acc. prob=0.77]
warmup: 14%|█▎ | 136/1000 [00:36<00:57, 15.12it/s, 127 steps of size 7.63e-02. acc. prob=0.77]
warmup: 14%|█▍ | 140/1000 [00:36<00:44, 19.53it/s, 63 steps of size 1.30e-01. acc. prob=0.77]
warmup: 14%|█▍ | 142/1000 [00:36<00:50, 16.96it/s, 255 steps of size 2.99e-02. acc. prob=0.77]
warmup: 14%|█▍ | 144/1000 [00:36<00:53, 15.89it/s, 127 steps of size 7.56e-02. acc. prob=0.77]
warmup: 15%|█▍ | 148/1000 [00:37<00:42, 19.99it/s, 63 steps of size 4.69e-02. acc. prob=0.77]
warmup: 15%|█▌ | 150/1000 [00:37<00:43, 19.54it/s, 63 steps of size 8.16e-02. acc. prob=0.77]
warmup: 16%|█▌ | 155/1000 [00:37<00:32, 26.23it/s, 63 steps of size 8.54e-02. acc. prob=0.77]
warmup: 16%|█▌ | 158/1000 [00:37<00:32, 26.31it/s, 63 steps of size 1.43e-01. acc. prob=0.77]
warmup: 16%|█▌ | 162/1000 [00:37<00:36, 23.14it/s, 255 steps of size 3.12e-02. acc. prob=0.77]
warmup: 16%|█▋ | 164/1000 [00:37<00:38, 21.97it/s, 63 steps of size 1.07e-01. acc. prob=0.77]
warmup: 17%|█▋ | 166/1000 [00:37<00:39, 20.88it/s, 127 steps of size 6.12e-02. acc. prob=0.77]
warmup: 17%|█▋ | 170/1000 [00:38<00:42, 19.60it/s, 255 steps of size 3.14e-02. acc. prob=0.77]
warmup: 17%|█▋ | 172/1000 [00:38<00:43, 19.22it/s, 63 steps of size 1.00e-01. acc. prob=0.77]
warmup: 18%|█▊ | 176/1000 [00:38<00:35, 23.36it/s, 31 steps of size 1.27e-01. acc. prob=0.77]
warmup: 18%|█▊ | 180/1000 [00:38<00:32, 25.51it/s, 63 steps of size 1.08e-01. acc. prob=0.77]
warmup: 18%|█▊ | 184/1000 [00:38<00:28, 28.17it/s, 31 steps of size 1.23e-01. acc. prob=0.78]
warmup: 19%|█▊ | 187/1000 [00:38<00:30, 26.71it/s, 63 steps of size 9.70e-02. acc. prob=0.77]
warmup: 19%|█▉ | 190/1000 [00:38<00:30, 26.77it/s, 63 steps of size 4.87e-02. acc. prob=0.77]
warmup: 19%|█▉ | 192/1000 [00:38<00:33, 24.41it/s, 127 steps of size 1.19e-01. acc. prob=0.78]
warmup: 20%|█▉ | 197/1000 [00:38<00:26, 30.47it/s, 31 steps of size 6.43e-02. acc. prob=0.77]
warmup: 20%|██ | 200/1000 [00:39<00:26, 29.91it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 20%|██ | 205/1000 [00:39<00:26, 30.47it/s, 127 steps of size 4.79e-02. acc. prob=0.78]
warmup: 21%|██ | 207/1000 [00:39<00:29, 27.27it/s, 63 steps of size 8.98e-02. acc. prob=0.78]
warmup: 21%|██ | 210/1000 [00:39<00:30, 26.23it/s, 127 steps of size 5.79e-02. acc. prob=0.78]
warmup: 21%|██ | 212/1000 [00:39<00:32, 24.16it/s, 63 steps of size 1.17e-01. acc. prob=0.78]
warmup: 22%|██▏ | 218/1000 [00:39<00:27, 28.35it/s, 127 steps of size 5.09e-02. acc. prob=0.78]
warmup: 22%|██▏ | 221/1000 [00:39<00:27, 28.02it/s, 63 steps of size 1.07e-01. acc. prob=0.78]
warmup: 23%|██▎ | 227/1000 [00:40<00:25, 30.65it/s, 127 steps of size 7.04e-02. acc. prob=0.78]
warmup: 23%|██▎ | 231/1000 [00:40<00:24, 30.94it/s, 63 steps of size 7.16e-02. acc. prob=0.78]
warmup: 23%|██▎ | 234/1000 [00:40<00:25, 30.13it/s, 63 steps of size 6.38e-02. acc. prob=0.78]
warmup: 24%|██▎ | 237/1000 [00:40<00:25, 29.45it/s, 63 steps of size 1.26e-01. acc. prob=0.78]
warmup: 24%|██▍ | 241/1000 [00:40<00:24, 31.56it/s, 63 steps of size 1.40e-01. acc. prob=0.78]
warmup: 24%|██▍ | 244/1000 [00:40<00:24, 30.54it/s, 127 steps of size 7.13e-02. acc. prob=0.78]
warmup: 25%|██▍ | 248/1000 [00:40<00:23, 32.55it/s, 31 steps of size 1.19e-01. acc. prob=0.78]
warmup: 25%|██▌ | 254/1000 [00:40<00:19, 37.47it/s, 63 steps of size 9.01e-02. acc. prob=0.78]
warmup: 26%|██▌ | 257/1000 [00:40<00:22, 33.09it/s, 127 steps of size 9.21e-02. acc. prob=0.78]
warmup: 26%|██▌ | 262/1000 [00:41<00:22, 33.49it/s, 127 steps of size 7.84e-02. acc. prob=0.78]
warmup: 27%|██▋ | 267/1000 [00:41<00:26, 27.96it/s, 255 steps of size 3.22e-02. acc. prob=0.78]
warmup: 27%|██▋ | 269/1000 [00:41<00:28, 25.88it/s, 63 steps of size 9.03e-02. acc. prob=0.78]
warmup: 27%|██▋ | 272/1000 [00:41<00:33, 22.06it/s, 255 steps of size 4.48e-02. acc. prob=0.78]
warmup: 27%|██▋ | 274/1000 [00:41<00:34, 21.30it/s, 63 steps of size 1.31e-01. acc. prob=0.78]
warmup: 28%|██▊ | 278/1000 [00:41<00:31, 22.92it/s, 127 steps of size 8.42e-02. acc. prob=0.78]
warmup: 28%|██▊ | 282/1000 [00:42<00:28, 25.20it/s, 127 steps of size 5.03e-02. acc. prob=0.78]
warmup: 28%|██▊ | 284/1000 [00:42<00:30, 23.46it/s, 63 steps of size 1.40e-01. acc. prob=0.78]
warmup: 29%|██▉ | 288/1000 [00:42<00:31, 22.82it/s, 255 steps of size 4.52e-02. acc. prob=0.78]
warmup: 29%|██▉ | 290/1000 [00:42<00:32, 21.74it/s, 63 steps of size 1.08e-01. acc. prob=0.78]
warmup: 29%|██▉ | 293/1000 [00:42<00:30, 23.36it/s, 63 steps of size 9.30e-02. acc. prob=0.78]
warmup: 30%|██▉ | 296/1000 [00:42<00:28, 24.57it/s, 63 steps of size 6.85e-02. acc. prob=0.78]
warmup: 30%|███ | 300/1000 [00:42<00:27, 25.62it/s, 127 steps of size 5.29e-02. acc. prob=0.78]
warmup: 30%|███ | 303/1000 [00:42<00:26, 26.26it/s, 63 steps of size 1.03e-01. acc. prob=0.78]
warmup: 31%|███ | 307/1000 [00:42<00:24, 28.76it/s, 127 steps of size 5.15e-02. acc. prob=0.78]
warmup: 31%|███ | 309/1000 [00:43<00:26, 25.89it/s, 63 steps of size 1.06e-01. acc. prob=0.78]
warmup: 31%|███ | 312/1000 [00:43<00:29, 23.15it/s, 127 steps of size 5.45e-02. acc. prob=0.78]
warmup: 32%|███▏ | 315/1000 [00:43<00:28, 24.43it/s, 63 steps of size 8.35e-02. acc. prob=0.78]
warmup: 32%|███▏ | 319/1000 [00:43<00:24, 27.97it/s, 31 steps of size 3.08e-02. acc. prob=0.78]
warmup: 32%|███▏ | 321/1000 [00:43<00:26, 25.32it/s, 63 steps of size 6.94e-02. acc. prob=0.78]
warmup: 32%|███▏ | 324/1000 [00:43<00:25, 26.05it/s, 63 steps of size 1.49e-01. acc. prob=0.78]
warmup: 33%|███▎ | 328/1000 [00:43<00:24, 27.99it/s, 63 steps of size 8.85e-02. acc. prob=0.78]
warmup: 33%|███▎ | 331/1000 [00:43<00:23, 27.96it/s, 63 steps of size 4.44e-02. acc. prob=0.78]
warmup: 33%|███▎ | 333/1000 [00:44<00:26, 25.27it/s, 63 steps of size 9.76e-02. acc. prob=0.78]
warmup: 34%|███▎ | 337/1000 [00:44<00:24, 27.39it/s, 63 steps of size 8.51e-02. acc. prob=0.78]
warmup: 34%|███▍ | 342/1000 [00:44<00:20, 32.78it/s, 31 steps of size 4.19e-02. acc. prob=0.78]
warmup: 34%|███▍ | 344/1000 [00:44<00:22, 28.53it/s, 63 steps of size 8.49e-02. acc. prob=0.78]
warmup: 35%|███▍ | 348/1000 [00:44<00:22, 29.59it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 35%|███▌ | 352/1000 [00:44<00:21, 30.36it/s, 63 steps of size 1.05e-01. acc. prob=0.78]
warmup: 36%|███▌ | 356/1000 [00:44<00:19, 32.26it/s, 63 steps of size 1.01e-01. acc. prob=0.78]
warmup: 36%|███▌ | 360/1000 [00:44<00:19, 33.07it/s, 127 steps of size 6.96e-02. acc. prob=0.78]
warmup: 36%|███▋ | 363/1000 [00:44<00:20, 31.59it/s, 63 steps of size 7.27e-02. acc. prob=0.78]
warmup: 37%|███▋ | 367/1000 [00:45<00:19, 33.22it/s, 31 steps of size 1.43e-01. acc. prob=0.78]
warmup: 37%|███▋ | 370/1000 [00:45<00:22, 27.48it/s, 127 steps of size 7.37e-02. acc. prob=0.78]
warmup: 37%|███▋ | 374/1000 [00:45<00:22, 28.30it/s, 63 steps of size 9.39e-02. acc. prob=0.78]
warmup: 38%|███▊ | 378/1000 [00:45<00:21, 28.97it/s, 63 steps of size 1.14e-01. acc. prob=0.78]
warmup: 38%|███▊ | 382/1000 [00:45<00:19, 31.01it/s, 63 steps of size 1.01e-01. acc. prob=0.78]
warmup: 39%|███▊ | 387/1000 [00:45<00:19, 32.06it/s, 127 steps of size 4.94e-02. acc. prob=0.78]
warmup: 39%|███▉ | 390/1000 [00:45<00:19, 30.74it/s, 63 steps of size 6.84e-02. acc. prob=0.78]
warmup: 40%|███▉ | 395/1000 [00:45<00:17, 34.95it/s, 31 steps of size 1.46e-01. acc. prob=0.78]
warmup: 40%|███▉ | 399/1000 [00:46<00:16, 35.49it/s, 31 steps of size 7.83e-02. acc. prob=0.78]
warmup: 40%|████ | 402/1000 [00:46<00:17, 33.44it/s, 63 steps of size 7.81e-02. acc. prob=0.78]
warmup: 41%|████ | 406/1000 [00:46<00:17, 33.09it/s, 63 steps of size 9.63e-02. acc. prob=0.78]
warmup: 41%|████ | 410/1000 [00:46<00:18, 31.33it/s, 127 steps of size 6.40e-02. acc. prob=0.78]
warmup: 41%|████▏ | 414/1000 [00:46<00:18, 31.48it/s, 63 steps of size 7.53e-02. acc. prob=0.78]
warmup: 42%|████▏ | 417/1000 [00:46<00:19, 30.41it/s, 63 steps of size 1.09e-01. acc. prob=0.78]
warmup: 42%|████▏ | 420/1000 [00:46<00:19, 29.41it/s, 63 steps of size 8.95e-02. acc. prob=0.78]
warmup: 42%|████▎ | 425/1000 [00:46<00:18, 30.70it/s, 127 steps of size 7.59e-02. acc. prob=0.78]
warmup: 43%|████▎ | 429/1000 [00:47<00:17, 32.31it/s, 31 steps of size 1.07e-01. acc. prob=0.78]
warmup: 43%|████▎ | 434/1000 [00:47<00:15, 36.00it/s, 31 steps of size 9.99e-02. acc. prob=0.78]
warmup: 44%|████▍ | 439/1000 [00:47<00:15, 36.96it/s, 63 steps of size 8.89e-02. acc. prob=0.78]
warmup: 44%|████▍ | 442/1000 [00:47<00:16, 34.29it/s, 63 steps of size 1.10e-01. acc. prob=0.78]
warmup: 45%|████▍ | 447/1000 [00:47<00:14, 37.20it/s, 31 steps of size 8.56e-02. acc. prob=0.78]
warmup: 45%|████▌ | 451/1000 [00:47<00:14, 36.99it/s, 63 steps of size 7.19e-01. acc. prob=0.78]
warmup: 46%|████▌ | 456/1000 [00:47<00:13, 39.50it/s, 31 steps of size 1.27e-01. acc. prob=0.78]
warmup: 46%|████▌ | 460/1000 [00:47<00:14, 36.15it/s, 63 steps of size 1.33e-01. acc. prob=0.78]
warmup: 46%|████▋ | 464/1000 [00:47<00:15, 34.48it/s, 63 steps of size 1.02e-01. acc. prob=0.78]
warmup: 47%|████▋ | 469/1000 [00:48<00:17, 30.38it/s, 255 steps of size 4.12e-02. acc. prob=0.78]
warmup: 47%|████▋ | 471/1000 [00:48<00:19, 27.35it/s, 63 steps of size 1.09e-01. acc. prob=0.78]
warmup: 48%|████▊ | 475/1000 [00:48<00:19, 27.41it/s, 127 steps of size 7.52e-02. acc. prob=0.78]
warmup: 48%|████▊ | 479/1000 [00:48<00:22, 23.64it/s, 255 steps of size 4.01e-02. acc. prob=0.78]
warmup: 48%|████▊ | 483/1000 [00:48<00:19, 26.56it/s, 31 steps of size 1.99e-01. acc. prob=0.78]
warmup: 49%|████▊ | 486/1000 [00:48<00:19, 26.07it/s, 63 steps of size 9.96e-02. acc. prob=0.78]
warmup: 49%|████▉ | 490/1000 [00:49<00:17, 28.63it/s, 31 steps of size 1.59e-01. acc. prob=0.78]
warmup: 50%|████▉ | 495/1000 [00:49<00:16, 31.53it/s, 63 steps of size 1.29e-01. acc. prob=0.78]
warmup: 50%|█████ | 500/1000 [00:49<00:14, 35.29it/s, 31 steps of size 9.98e-02. acc. prob=0.78]
sample: 50%|█████ | 504/1000 [00:49<00:14, 34.32it/s, 63 steps of size 9.98e-02. acc. prob=0.90]
sample: 51%|█████ | 509/1000 [00:49<00:13, 37.43it/s, 63 steps of size 9.98e-02. acc. prob=0.92]
sample: 51%|█████▏ | 514/1000 [00:49<00:12, 39.67it/s, 31 steps of size 9.98e-02. acc. prob=0.92]
sample: 52%|█████▏ | 519/1000 [00:49<00:11, 41.41it/s, 31 steps of size 9.98e-02. acc. prob=0.91]
sample: 52%|█████▏ | 523/1000 [00:49<00:12, 39.64it/s, 63 steps of size 9.98e-02. acc. prob=0.92]
sample: 53%|█████▎ | 526/1000 [00:49<00:13, 36.17it/s, 63 steps of size 9.98e-02. acc. prob=0.92]
sample: 53%|█████▎ | 529/1000 [00:50<00:14, 31.85it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 53%|█████▎ | 532/1000 [00:50<00:15, 30.05it/s, 63 steps of size 9.98e-02. acc. prob=0.90]
sample: 54%|█████▎ | 536/1000 [00:50<00:14, 31.47it/s, 31 steps of size 9.98e-02. acc. prob=0.90]
sample: 54%|█████▍ | 539/1000 [00:50<00:15, 29.82it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 54%|█████▍ | 542/1000 [00:50<00:15, 28.79it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 55%|█████▍ | 546/1000 [00:50<00:15, 29.40it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 55%|█████▌ | 550/1000 [00:50<00:15, 29.95it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 55%|█████▌ | 552/1000 [00:50<00:16, 26.62it/s, 127 steps of size 9.98e-02. acc. prob=0.88]
sample: 56%|█████▌ | 556/1000 [00:51<00:15, 27.98it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 56%|█████▌ | 559/1000 [00:51<00:15, 27.70it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 56%|█████▌ | 562/1000 [00:51<00:15, 27.51it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 56%|█████▋ | 565/1000 [00:51<00:15, 27.29it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 57%|█████▋ | 569/1000 [00:51<00:14, 29.86it/s, 31 steps of size 9.98e-02. acc. prob=0.89]
sample: 57%|█████▋ | 572/1000 [00:51<00:14, 29.06it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 57%|█████▊ | 575/1000 [00:51<00:14, 28.57it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 58%|█████▊ | 579/1000 [00:51<00:14, 29.46it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 58%|█████▊ | 582/1000 [00:51<00:14, 28.75it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 58%|█████▊ | 585/1000 [00:52<00:14, 28.19it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 59%|█████▉ | 589/1000 [00:52<00:14, 29.03it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 59%|█████▉ | 593/1000 [00:52<00:13, 29.62it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 60%|█████▉ | 596/1000 [00:52<00:13, 28.89it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 60%|██████ | 600/1000 [00:52<00:13, 29.40it/s, 63 steps of size 9.98e-02. acc. prob=0.89]
sample: 60%|██████ | 604/1000 [00:52<00:13, 28.61it/s, 127 steps of size 9.98e-02. acc. prob=0.89]
sample: 61%|██████ | 607/1000 [00:52<00:13, 28.14it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 61%|██████ | 610/1000 [00:52<00:13, 27.86it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 61%|██████▏ | 613/1000 [00:53<00:14, 27.62it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 62%|██████▏ | 617/1000 [00:53<00:13, 28.74it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 62%|██████▏ | 621/1000 [00:53<00:12, 29.49it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 62%|██████▏ | 624/1000 [00:53<00:13, 28.80it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 63%|██████▎ | 627/1000 [00:53<00:13, 28.20it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 63%|██████▎ | 630/1000 [00:53<00:13, 26.64it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 63%|██████▎ | 633/1000 [00:53<00:14, 25.82it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 64%|██████▎ | 637/1000 [00:53<00:13, 26.98it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 64%|██████▍ | 640/1000 [00:54<00:13, 26.67it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 64%|██████▍ | 644/1000 [00:54<00:12, 27.47it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 65%|██████▍ | 648/1000 [00:54<00:12, 28.05it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 65%|██████▌ | 652/1000 [00:54<00:12, 28.71it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 66%|██████▌ | 656/1000 [00:54<00:11, 29.40it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 66%|██████▌ | 660/1000 [00:54<00:11, 30.89it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 66%|██████▋ | 664/1000 [00:54<00:10, 32.26it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 67%|██████▋ | 667/1000 [00:54<00:10, 30.74it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 67%|██████▋ | 670/1000 [00:54<00:11, 29.67it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 67%|██████▋ | 674/1000 [00:55<00:10, 30.11it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 68%|██████▊ | 677/1000 [00:55<00:11, 29.12it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 68%|██████▊ | 681/1000 [00:55<00:10, 31.10it/s, 31 steps of size 9.98e-02. acc. prob=0.87]
sample: 68%|██████▊ | 685/1000 [00:55<00:10, 31.20it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 69%|██████▉ | 688/1000 [00:55<00:10, 30.16it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 69%|██████▉ | 691/1000 [00:55<00:10, 29.22it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 69%|██████▉ | 694/1000 [00:55<00:10, 28.61it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 70%|██████▉ | 698/1000 [00:55<00:09, 30.90it/s, 31 steps of size 9.98e-02. acc. prob=0.87]
sample: 70%|███████ | 702/1000 [00:56<00:09, 32.34it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 70%|███████ | 705/1000 [00:56<00:09, 29.95it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 71%|███████ | 708/1000 [00:56<00:10, 28.63it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 71%|███████ | 711/1000 [00:56<00:10, 28.14it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 71%|███████▏ | 714/1000 [00:56<00:10, 27.86it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 72%|███████▏ | 717/1000 [00:56<00:10, 27.56it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 72%|███████▏ | 721/1000 [00:56<00:09, 30.18it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 72%|███████▎ | 725/1000 [00:56<00:09, 30.53it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 73%|███████▎ | 729/1000 [00:56<00:08, 32.16it/s, 63 steps of size 9.98e-02. acc. prob=0.87]
sample: 73%|███████▎ | 733/1000 [00:57<00:08, 31.04it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 74%|███████▎ | 736/1000 [00:57<00:08, 30.60it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 74%|███████▍ | 741/1000 [00:57<00:07, 32.90it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 74%|███████▍ | 744/1000 [00:57<00:08, 30.84it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 75%|███████▍ | 749/1000 [00:57<00:07, 34.53it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 75%|███████▌ | 753/1000 [00:57<00:07, 34.68it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 76%|███████▌ | 757/1000 [00:57<00:06, 34.76it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 76%|███████▌ | 762/1000 [00:57<00:06, 37.48it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 77%|███████▋ | 766/1000 [00:58<00:06, 34.79it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 77%|███████▋ | 771/1000 [00:58<00:06, 37.31it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 78%|███████▊ | 776/1000 [00:58<00:05, 39.23it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 78%|███████▊ | 779/1000 [00:58<00:06, 35.42it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 78%|███████▊ | 784/1000 [00:58<00:05, 37.88it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 79%|███████▉ | 788/1000 [00:58<00:06, 35.18it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 79%|███████▉ | 791/1000 [00:58<00:06, 32.52it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 80%|███████▉ | 795/1000 [00:58<00:06, 31.64it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 80%|███████▉ | 799/1000 [00:58<00:06, 32.51it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 80%|████████ | 803/1000 [00:59<00:06, 31.70it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 81%|████████ | 806/1000 [00:59<00:06, 29.99it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 81%|████████ | 809/1000 [00:59<00:06, 27.50it/s, 95 steps of size 9.98e-02. acc. prob=0.88]
sample: 81%|████████▏ | 813/1000 [00:59<00:06, 28.12it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 82%|████████▏ | 817/1000 [00:59<00:06, 28.69it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 82%|████████▏ | 822/1000 [00:59<00:05, 31.33it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 82%|████████▎ | 825/1000 [00:59<00:05, 29.86it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 83%|████████▎ | 828/1000 [00:59<00:05, 28.86it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 83%|████████▎ | 831/1000 [01:00<00:06, 28.06it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 83%|████████▎ | 834/1000 [01:00<00:06, 27.58it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 84%|████████▎ | 837/1000 [01:00<00:06, 27.14it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 84%|████████▍ | 841/1000 [01:00<00:05, 27.96it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 84%|████████▍ | 845/1000 [01:00<00:05, 28.86it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 85%|████████▍ | 849/1000 [01:00<00:05, 29.45it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 85%|████████▌ | 852/1000 [01:00<00:05, 28.88it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 86%|████████▌ | 856/1000 [01:00<00:04, 29.29it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 86%|████████▌ | 860/1000 [01:01<00:04, 29.53it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 86%|████████▋ | 863/1000 [01:01<00:04, 28.97it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 87%|████████▋ | 866/1000 [01:01<00:04, 28.44it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 87%|████████▋ | 870/1000 [01:01<00:04, 30.62it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 87%|████████▋ | 873/1000 [01:01<00:04, 29.51it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 88%|████████▊ | 877/1000 [01:01<00:04, 29.89it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 88%|████████▊ | 880/1000 [01:01<00:04, 29.05it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 88%|████████▊ | 884/1000 [01:01<00:03, 29.54it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 89%|████████▊ | 887/1000 [01:02<00:03, 28.81it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 89%|████████▉ | 890/1000 [01:02<00:03, 28.36it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 89%|████████▉ | 894/1000 [01:02<00:03, 30.59it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 90%|████████▉ | 898/1000 [01:02<00:03, 30.66it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 90%|█████████ | 901/1000 [01:02<00:03, 29.65it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 90%|█████████ | 905/1000 [01:02<00:03, 30.00it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 91%|█████████ | 910/1000 [01:02<00:02, 32.70it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 91%|█████████▏| 913/1000 [01:02<00:02, 30.94it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 92%|█████████▏| 916/1000 [01:02<00:02, 29.98it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 92%|█████████▏| 920/1000 [01:03<00:02, 31.70it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 92%|█████████▏| 923/1000 [01:03<00:02, 30.43it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 93%|█████████▎| 927/1000 [01:03<00:02, 32.14it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 93%|█████████▎| 931/1000 [01:03<00:02, 33.53it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 94%|█████████▎| 936/1000 [01:03<00:01, 35.33it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 94%|█████████▍| 939/1000 [01:03<00:01, 33.04it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 94%|█████████▍| 943/1000 [01:03<00:01, 32.46it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 95%|█████████▍| 947/1000 [01:03<00:01, 33.55it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 95%|█████████▌| 952/1000 [01:04<00:01, 35.13it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 96%|█████████▌| 956/1000 [01:04<00:01, 32.50it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 96%|█████████▌| 959/1000 [01:04<00:01, 31.44it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 96%|█████████▌| 962/1000 [01:04<00:01, 31.06it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 96%|█████████▋| 965/1000 [01:04<00:01, 28.91it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 97%|█████████▋| 969/1000 [01:04<00:01, 28.90it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 97%|█████████▋| 972/1000 [01:04<00:00, 28.07it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 98%|█████████▊| 975/1000 [01:04<00:00, 27.28it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 98%|█████████▊| 978/1000 [01:04<00:00, 26.78it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 98%|█████████▊| 982/1000 [01:05<00:00, 27.46it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 99%|█████████▊| 986/1000 [01:05<00:00, 29.69it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
sample: 99%|█████████▉| 989/1000 [01:05<00:00, 28.89it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 99%|█████████▉| 993/1000 [01:05<00:00, 30.87it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 100%|█████████▉| 996/1000 [01:05<00:00, 29.70it/s, 63 steps of size 9.98e-02. acc. prob=0.88]
sample: 100%|██████████| 1000/1000 [01:05<00:00, 15.23it/s, 31 steps of size 9.98e-02. acc. prob=0.88]
There was 1 divergence after tuning. Increase `target_accept` or reparameterize.
We recommend running at least 4 chains for robust computation of convergence diagnostics
The rhat statistic is larger than 1.01 for some parameters. This indicates problems during sampling. See https://arxiv.org/abs/1903.08008 for details
The effective sample size per chain is smaller than 100 for some parameters. A higher number is needed for reliable rhat and ess computation. See https://arxiv.org/abs/1903.08008 for details
In [33]:
Copied!
#### posterior statistics of non-centered model
az.summary(
samples_model_reg_v_ex4_A2, var_names=["~_offset", "~_id"], filter_vars="like"
)
#### posterior statistics of non-centered model
az.summary(
samples_model_reg_v_ex4_A2, var_names=["~_offset", "~_id"], filter_vars="like"
)
Out[33]:
| mean | sd | eti89_lb | eti89_ub | ess_bulk | ess_tail | r_hat | mcse_mean | mcse_sd | |
|---|---|---|---|---|---|---|---|---|---|
| v_zGPe | 0.375 | 0.047 | 0.31 | 0.45 | 435 | 478 | 1.01 | 0.0023 | 0.0021 |
| z | 0.4979 | 0.0062 | 0.49 | 0.51 | 2266 | 1120 | 1.00 | 0.00013 | 9.2e-05 |
| v_sevScore | -0.12 | 0.21 | -0.45 | 0.2 | 758 | 759 | 1.01 | 0.0075 | 0.0058 |
| v_zGPe:sevScore | -0.502 | 0.095 | -0.66 | -0.36 | 491 | 608 | 1.00 | 0.0043 | 0.0038 |
| v_zSTN | 0.867 | 0.058 | 0.78 | 0.96 | 721 | 866 | 1.00 | 0.0023 | 0.0022 |
| v_Intercept | 0.507 | 0.109 | 0.34 | 0.68 | 625 | 679 | 1.00 | 0.0044 | 0.0038 |
| v_zSTN:sevScore | -0.66 | 0.12 | -0.84 | -0.48 | 568 | 580 | 1.00 | 0.0053 | 0.0055 |
| t | 0.5177 | 0.0064 | 0.51 | 0.53 | 1959 | 1199 | 1.00 | 0.00014 | 0.00011 |
| a | 1.4981 | 0.0165 | 1.5 | 1.5 | 2272 | 1074 | 1.00 | 0.00035 | 0.00025 |
Comparison¶
In [34]:
Copied!
az.compare(
{"Model 1": samples_model_reg_v_ex4_A1, "Model 2": samples_model_reg_v_ex4_A2}
)
az.compare(
{"Model 1": samples_model_reg_v_ex4_A1, "Model 2": samples_model_reg_v_ex4_A2}
)
Out[34]:
| rank | elpd | p | elpd_diff | weight | se | dse | warning | |
|---|---|---|---|---|---|---|---|---|
| Model 1 | 0 | -5500.0 | 24.9 | 0.0 | 1.0 | 90.0 | 0.0 | False |
| Model 2 | 1 | -5500.0 | 34.6 | -6.0 | 0.0 | 90.0 | 2.6 | False |