# %%
from agoge import InferenceWorker
import threading
import torch
import mido
import time
import numpy as np
from tqdm import tqdm
import jack
from matplotlib import pyplot as plt
from itertools import cycle
from numpy import array
worker = InferenceWorker('hasty-copper-dogfish', 'dx7-vae', with_data=True)
float32 = 'float32'
model = worker.model
# data = worker.dataset
# loader = data.loaders.test
n_samples = 32
n_latents = 8
# loader.batch_sampler.batch_size = n_samples
# randoms = torch.cat([model.generate(torch.randn(2**11, 8)).logits.argmax(-1) for _ in tqdm(range(2**5))])
# %%
from matplotlib import pyplot as plt
import torch
rand = torch.rand(100)
randn = torch.randn(100)
plt.scatter(rand, torch.sigmoid(rand) + 0.5)
# %%
from agoge import InferenceWorker
import threading
import torch
import time
import numpy as np
from tqdm import tqdm
import jack
from matplotlib import pyplot as plt
from itertools import cycle
worker = InferenceWorker(
'~/agoge/artifacts/dx7-vae/hasty-copper-dogfish_0_2020-05-06_10-46-27o654hmde/checkpoint_204/model.box',
with_data=True)
model = worker.model
data = worker.dataset
loader = data.loaders.test
n_samples = 32
n_latents = 8
loader.batch_sampler.batch_size = n_samples
features_all = []
features_half = []
for x in map(lambda x: x['X'], tqdm(loader)):
q = model.features(x)
features_all += [(q.mean.numpy(), q.stddev.numpy())]
# features_half += [model.features(x, torch.rand_like(x.float())>torch.linspace(0, 1, 32).unsqueeze(-1)).mean.numpy()]
mus, vars = map(np.concatenate, zip(*features_all))
print(mus.mean(0), vars.mean(0))
# for item in loader:
# %%
from agoge import InferenceWorker
import torch
from tqdm import tqdm
from matplotlib import pyplot as plt
worker = InferenceWorker(
'/home/nintorac/agoge/artifacts/craggy-goldenrod-catfish_0_2020-04-28_02-22-57m8eftq1b/checkpoint_410/model.box',
with_data=True)
model = worker.model
data = worker.dataset
loader = data.loaders.test
n_samples = 32
loader.batch_sampler.batch_size = n_samples
# %%
# batch = next(iter(loader))['x']
# X_a = torch.rand_like(batch.float()) > torch.linspace(0, 1, n_samples).unsqueeze(-1)
# logits = model.generate(batch, X_a)
# # %%
# from matplotlib import pyplot as plt
# plt.imshow(X_a)
# # %%
# plt.scatter(torch.arange(n_samples), logits.log_prob(batch).mean(-1))
# %%
from agoge import InferenceWorker
import threading
import torch
import time
import numpy as np
from tqdm import tqdm
import jack
from matplotlib import pyplot as plt
from itertools import cycle
worker = InferenceWorker(
'/home/nintorac/agoge/artifacts/bluesy-chestnut-forest_0_2020-04-30_11-11-00x02v59be/checkpoint_220/model.box',
with_data=True)
model = worker.model
data = worker.dataset
loader = data.loaders.test
n_samples = 32
n_latents = 8
loader.batch_sampler.batch_size = n_samples
features_all = []
features_half = []
for x in map(lambda x: x['x'], tqdm(loader)):
q = model.features(x, torch.ones_like(x.float()).bool())
features_all += [(q.mean.numpy(), q.stddev.numpy())]
# features_half += [model.features(x, torch.rand_like(x.float())>torch.linspace(0, 1, 32).unsqueeze(-1)).mean.numpy()]
# for item in loader:
# %%
# %%
from agoge import InferenceWorker
import torch
from tqdm import tqdm
from matplotlib import pyplot as plt
worker = InferenceWorker(
'/home/nintorac/agoge/artifacts/Worker/messy-firebrick-barracuda_0_2020-04-14_08-59-40ryp93n44/checkpoint_4/model.box',
with_data=True)
model = worker.model
data = worker.dataset
loader = data.loaders.test
Xs = []
features = []
for X in tqdm(loader):
Xs += [X['x']]
features += [model.features(X['x'])]
features = torch.cat(features).flatten(-2, -1)
X = torch.cat(Xs)
#%%
### --------TSNE-----------
import numpy as np
from sklearn.manifold import TSNE
X_embedded = TSNE(n_components=2).fit_transform(features)
X_embedded.shape
#%%
plt.figure(figsize=(20, 30))
plt.scatter(*zip(*X_embedded), linewidths=0.1)
# %%
from agoge import InferenceWorker
import threading
import torch
import mido
import time
import numpy as np
from tqdm import tqdm
import jack
from matplotlib import pyplot as plt
from itertools import cycle
worker = InferenceWorker('~/agoge/artifacts/dx7-nsp/leaky-burgundy-coati.box',
with_data=True)
model = worker.model
data = worker.dataset
loader = data.loaders.test
n_samples = 32
n_latents = 8
loader.batch_sampler.batch_size = n_samples
from uuid import uuid4 as uuid
uuid = lambda: hex(uuid)
# self._event.set()
client = jack.Client('DX7Parameteriser')
port = client.midi_outports.register('output')
inport = client.midi_inports.register('input')
event = threading.Event()
fs = None # sampling rate
# %%
from agoge import InferenceWorker
import threading
import torch
import mido
import time
import numpy as np
from tqdm import tqdm
import jack
from matplotlib import pyplot as plt
from itertools import cycle
worker = InferenceWorker(
'/home/nintorac/agoge/artifacts/squeaky-green-mist_0_2020-04-30_12-50-391i661hyj/checkpoint_100/model.box',
with_data=True)
model = worker.model
data = worker.dataset
loader = data.loaders.test
n_samples = 32
n_latents = 8
loader.batch_sampler.batch_size = n_samples
from uuid import uuid4 as uuid
uuid = lambda: hex(uuid)
# self._event.set()
client = jack.Client('DX7Parameteriser')
port = client.midi_outports.register('output')
inport = client.midi_inports.register('input')
event = threading.Event()