Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[FIX] Minor bugs #1158

Merged
merged 5 commits into from
Oct 1, 2024
Merged

[FIX] Minor bugs #1158

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
536e8f6
small_speed_improvements
elephaint Sep 26, 2024
e08ed44
add_back_item()
elephaint Sep 27, 2024
8d092b5
pad_float_not_int
elephaint Sep 30, 2024
4e5a0c5
Merge branch 'main' into fix/small_bugs
elephaint Sep 30, 2024
46921b4
Merge branch 'main' into fix/small_bugs
elephaint Oct 1, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions action_files/test_models/src/multivariate_models.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@
from neuralforecast.models.tsmixer import TSMixer
from neuralforecast.models.tsmixerx import TSMixerx
from neuralforecast.models.itransformer import iTransformer
from neuralforecast.models.stemgnn import StemGNN
# from neuralforecast.models.stemgnn import StemGNN
from neuralforecast.models.mlpmultivariate import MLPMultivariate
from neuralforecast.models.timemixer import TimeMixer

Expand All @@ -30,7 +30,7 @@ def main(dataset: str = 'multivariate', group: str = 'ETTm2') -> None:
TSMixer(h=horizon, n_series=7, input_size=2 * horizon, loss=MAE(), dropout=0.0, max_steps=1000, val_check_steps=500),
TSMixerx(h=horizon, n_series=7, input_size=2*horizon, loss=MAE(), dropout=0.0, max_steps=1000, val_check_steps=500),
iTransformer(h=horizon, n_series=7, input_size=2 * horizon, loss=MAE(), dropout=0.0, max_steps=1000, val_check_steps=500),
StemGNN(h=horizon, n_series=7, input_size=2*horizon, loss=MAE(), dropout_rate=0.0, max_steps=1000, val_check_steps=500),
# StemGNN(h=horizon, n_series=7, input_size=2*horizon, loss=MAE(), dropout_rate=0.0, max_steps=1000, val_check_steps=500),
MLPMultivariate(h=horizon, n_series=7, input_size=2*horizon, loss=MAE(), max_steps=1000, val_check_steps=500),
TimeMixer(h=horizon, n_series=7, input_size=2*horizon, loss=MAE(), dropout=0.0, max_steps=1000, val_check_steps=500)
]
Expand Down
2 changes: 1 addition & 1 deletion nbs/common.base_model.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -441,7 +441,7 @@
" if self.val_size == 0:\n",
" return\n",
" losses = torch.stack(self.validation_step_outputs)\n",
" avg_loss = losses.mean().item()\n",
" avg_loss = losses.mean().detach().item()\n",
" self.log(\n",
" \"ptl/val_loss\",\n",
" avg_loss,\n",
Expand Down
8 changes: 4 additions & 4 deletions nbs/common.base_multivariate.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,7 +131,7 @@
" self.h = h\n",
" self.input_size = input_size\n",
" self.n_series = n_series\n",
" self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0)\n",
" self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0.0)\n",
"\n",
" # Multivariate models do not support these loss functions yet.\n",
" unsupported_losses = (\n",
Expand Down Expand Up @@ -390,12 +390,12 @@
"\n",
" self.log(\n",
" 'train_loss',\n",
" loss.item(),\n",
" loss.detach().item(),\n",
" batch_size=outsample_y.size(0),\n",
" prog_bar=True,\n",
" on_epoch=True,\n",
" )\n",
" self.train_trajectories.append((self.global_step, loss.item()))\n",
" self.train_trajectories.append((self.global_step, loss.detach().item()))\n",
" return loss\n",
"\n",
" def validation_step(self, batch, batch_idx):\n",
Expand Down Expand Up @@ -440,7 +440,7 @@
"\n",
" self.log(\n",
" 'valid_loss',\n",
" valid_loss.item(),\n",
" valid_loss.detach().item(),\n",
" batch_size=outsample_y.size(0),\n",
" prog_bar=True,\n",
" on_epoch=True,\n",
Expand Down
10 changes: 5 additions & 5 deletions nbs/common.base_recurrent.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -137,7 +137,7 @@
" self.h = h\n",
" self.input_size = input_size\n",
" self.inference_input_size = inference_input_size\n",
" self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0)\n",
" self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0.0)\n",
"\n",
" unsupported_distributions = ['Bernoulli', 'ISQF']\n",
" if isinstance(self.loss, losses.DistributionLoss) and\\\n",
Expand Down Expand Up @@ -254,7 +254,7 @@
"\n",
" # Test size covers all data, pad left one timestep with zeros\n",
" if temporal.shape[-1] == self.test_size:\n",
" padder_left = nn.ConstantPad1d(padding=(1, 0), value=0)\n",
" padder_left = nn.ConstantPad1d(padding=(1, 0), value=0.0)\n",
" temporal = padder_left(temporal)\n",
"\n",
" # Parse batch\n",
Expand Down Expand Up @@ -365,12 +365,12 @@
"\n",
" self.log(\n",
" 'train_loss',\n",
" loss.item(),\n",
" loss.detach().item(),\n",
" batch_size=outsample_y.size(0),\n",
" prog_bar=True,\n",
" on_epoch=True,\n",
" )\n",
" self.train_trajectories.append((self.global_step, loss.item()))\n",
" self.train_trajectories.append((self.global_step, loss.detach().item()))\n",
" return loss\n",
"\n",
" def validation_step(self, batch, batch_idx):\n",
Expand Down Expand Up @@ -438,7 +438,7 @@
"\n",
" self.log(\n",
" 'valid_loss',\n",
" valid_loss.item(),\n",
" valid_loss.detach().item(),\n",
" batch_size=outsample_y.size(0),\n",
" prog_bar=True,\n",
" on_epoch=True,\n",
Expand Down
16 changes: 8 additions & 8 deletions nbs/common.base_windows.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -143,9 +143,9 @@
" self.windows_batch_size = windows_batch_size\n",
" self.start_padding_enabled = start_padding_enabled\n",
" if start_padding_enabled:\n",
" self.padder_train = nn.ConstantPad1d(padding=(self.input_size-1, self.h), value=0)\n",
" self.padder_train = nn.ConstantPad1d(padding=(self.input_size-1, self.h), value=0.0)\n",
" else:\n",
" self.padder_train = nn.ConstantPad1d(padding=(0, self.h), value=0)\n",
" self.padder_train = nn.ConstantPad1d(padding=(0, self.h), value=0.0)\n",
"\n",
" # Batch sizes\n",
" self.batch_size = batch_size\n",
Expand Down Expand Up @@ -265,7 +265,7 @@
" if step == 'predict':\n",
" initial_input = temporal.shape[-1] - self.test_size\n",
" if initial_input <= self.input_size: # There is not enough data to predict first timestamp\n",
" padder_left = nn.ConstantPad1d(padding=(self.input_size-initial_input, 0), value=0)\n",
" padder_left = nn.ConstantPad1d(padding=(self.input_size-initial_input, 0), value=0.0)\n",
" temporal = padder_left(temporal)\n",
" predict_step_size = self.predict_step_size\n",
" cutoff = - self.input_size - self.test_size\n",
Expand All @@ -280,11 +280,11 @@
" temporal = batch['temporal'][:, :, cutoff:]\n",
" if temporal.shape[-1] < window_size:\n",
" initial_input = temporal.shape[-1] - self.val_size\n",
" padder_left = nn.ConstantPad1d(padding=(self.input_size-initial_input, 0), value=0)\n",
" padder_left = nn.ConstantPad1d(padding=(self.input_size-initial_input, 0), value=0.0)\n",
" temporal = padder_left(temporal)\n",
"\n",
" if (step=='predict') and (self.test_size==0) and (len(self.futr_exog_list)==0):\n",
" padder_right = nn.ConstantPad1d(padding=(0, self.h), value=0)\n",
" padder_right = nn.ConstantPad1d(padding=(0, self.h), value=0.0)\n",
" temporal = padder_right(temporal)\n",
"\n",
" windows = temporal.unfold(dimension=-1,\n",
Expand Down Expand Up @@ -447,12 +447,12 @@
"\n",
" self.log(\n",
" 'train_loss',\n",
" loss.item(),\n",
" loss.detach().item(),\n",
" batch_size=outsample_y.size(0),\n",
" prog_bar=True,\n",
" on_epoch=True,\n",
" )\n",
" self.train_trajectories.append((self.global_step, loss.item()))\n",
" self.train_trajectories.append((self.global_step, loss.detach().item()))\n",
" return loss\n",
"\n",
" def _compute_valid_loss(self, outsample_y, output, outsample_mask, temporal_cols, y_idx):\n",
Expand Down Expand Up @@ -533,7 +533,7 @@
"\n",
" self.log(\n",
" 'valid_loss',\n",
" valid_loss.item(),\n",
" valid_loss.detach().item(),\n",
" batch_size=batch_size,\n",
" prog_bar=True,\n",
" on_epoch=True,\n",
Expand Down
6 changes: 2 additions & 4 deletions nbs/losses.pytorch.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,9 +105,7 @@
" Auxiliary funtion to handle divide by 0\n",
" \"\"\"\n",
" div = a / b\n",
" div[div != div] = 0.0\n",
" div[div == float('inf')] = 0.0\n",
" return div"
" return torch.nan_to_num(div, nan=0.0, posinf=0.0, neginf=0.0)"
]
},
{
Expand Down Expand Up @@ -1412,7 +1410,7 @@
" if (loc is not None) and (scale is not None):\n",
" mean = (mean * scale) + loc\n",
" tscale = (tscale + eps) * scale\n",
" df = 2.0 + F.softplus(df)\n",
" df = 3.0 + F.softplus(df)\n",
" return (df, mean, tscale)\n",
"\n",
"def normal_domain_map(input: torch.Tensor):\n",
Expand Down
2 changes: 1 addition & 1 deletion neuralforecast/common/_base_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -421,7 +421,7 @@ def on_validation_epoch_end(self):
if self.val_size == 0:
return
losses = torch.stack(self.validation_step_outputs)
avg_loss = losses.mean().item()
avg_loss = losses.mean().detach().item()
self.log(
"ptl/val_loss",
avg_loss,
Expand Down
8 changes: 4 additions & 4 deletions neuralforecast/common/_base_multivariate.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,7 +77,7 @@ def __init__(
self.h = h
self.input_size = input_size
self.n_series = n_series
self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0)
self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0.0)

# Multivariate models do not support these loss functions yet.
unsupported_losses = (
Expand Down Expand Up @@ -389,12 +389,12 @@ def training_step(self, batch, batch_idx):

self.log(
"train_loss",
loss.item(),
loss.detach().item(),
batch_size=outsample_y.size(0),
prog_bar=True,
on_epoch=True,
)
self.train_trajectories.append((self.global_step, loss.item()))
self.train_trajectories.append((self.global_step, loss.detach().item()))
return loss

def validation_step(self, batch, batch_idx):
Expand Down Expand Up @@ -456,7 +456,7 @@ def validation_step(self, batch, batch_idx):

self.log(
"valid_loss",
valid_loss.item(),
valid_loss.detach().item(),
batch_size=outsample_y.size(0),
prog_bar=True,
on_epoch=True,
Expand Down
10 changes: 5 additions & 5 deletions neuralforecast/common/_base_recurrent.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,7 +77,7 @@ def __init__(
self.h = h
self.input_size = input_size
self.inference_input_size = inference_input_size
self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0)
self.padder = nn.ConstantPad1d(padding=(0, self.h), value=0.0)

unsupported_distributions = ["Bernoulli", "ISQF"]
if (
Expand Down Expand Up @@ -210,7 +210,7 @@ def _create_windows(self, batch, step):

# Test size covers all data, pad left one timestep with zeros
if temporal.shape[-1] == self.test_size:
padder_left = nn.ConstantPad1d(padding=(1, 0), value=0)
padder_left = nn.ConstantPad1d(padding=(1, 0), value=0.0)
temporal = padder_left(temporal)

# Parse batch
Expand Down Expand Up @@ -349,12 +349,12 @@ def training_step(self, batch, batch_idx):

self.log(
"train_loss",
loss.item(),
loss.detach().item(),
batch_size=outsample_y.size(0),
prog_bar=True,
on_epoch=True,
)
self.train_trajectories.append((self.global_step, loss.item()))
self.train_trajectories.append((self.global_step, loss.detach().item()))
return loss

def validation_step(self, batch, batch_idx):
Expand Down Expand Up @@ -447,7 +447,7 @@ def validation_step(self, batch, batch_idx):

self.log(
"valid_loss",
valid_loss.item(),
valid_loss.detach().item(),
batch_size=outsample_y.size(0),
prog_bar=True,
on_epoch=True,
Expand Down
16 changes: 8 additions & 8 deletions neuralforecast/common/_base_windows.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,10 +83,10 @@ def __init__(
self.start_padding_enabled = start_padding_enabled
if start_padding_enabled:
self.padder_train = nn.ConstantPad1d(
padding=(self.input_size - 1, self.h), value=0
padding=(self.input_size - 1, self.h), value=0.0
)
else:
self.padder_train = nn.ConstantPad1d(padding=(0, self.h), value=0)
self.padder_train = nn.ConstantPad1d(padding=(0, self.h), value=0.0)

# Batch sizes
self.batch_size = batch_size
Expand Down Expand Up @@ -216,7 +216,7 @@ def _create_windows(self, batch, step, w_idxs=None):
initial_input <= self.input_size
): # There is not enough data to predict first timestamp
padder_left = nn.ConstantPad1d(
padding=(self.input_size - initial_input, 0), value=0
padding=(self.input_size - initial_input, 0), value=0.0
)
temporal = padder_left(temporal)
predict_step_size = self.predict_step_size
Expand All @@ -233,7 +233,7 @@ def _create_windows(self, batch, step, w_idxs=None):
if temporal.shape[-1] < window_size:
initial_input = temporal.shape[-1] - self.val_size
padder_left = nn.ConstantPad1d(
padding=(self.input_size - initial_input, 0), value=0
padding=(self.input_size - initial_input, 0), value=0.0
)
temporal = padder_left(temporal)

Expand All @@ -242,7 +242,7 @@ def _create_windows(self, batch, step, w_idxs=None):
and (self.test_size == 0)
and (len(self.futr_exog_list) == 0)
):
padder_right = nn.ConstantPad1d(padding=(0, self.h), value=0)
padder_right = nn.ConstantPad1d(padding=(0, self.h), value=0.0)
temporal = padder_right(temporal)

windows = temporal.unfold(
Expand Down Expand Up @@ -440,12 +440,12 @@ def training_step(self, batch, batch_idx):

self.log(
"train_loss",
loss.item(),
loss.detach().item(),
batch_size=outsample_y.size(0),
prog_bar=True,
on_epoch=True,
)
self.train_trajectories.append((self.global_step, loss.item()))
self.train_trajectories.append((self.global_step, loss.detach().item()))
return loss

def _compute_valid_loss(
Expand Down Expand Up @@ -551,7 +551,7 @@ def validation_step(self, batch, batch_idx):

self.log(
"valid_loss",
valid_loss.item(),
valid_loss.detach().item(),
batch_size=batch_size,
prog_bar=True,
on_epoch=True,
Expand Down
6 changes: 2 additions & 4 deletions neuralforecast/losses/pytorch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,9 +35,7 @@ def _divide_no_nan(a: torch.Tensor, b: torch.Tensor) -> torch.Tensor:
Auxiliary funtion to handle divide by 0
"""
div = a / b
div[div != div] = 0.0
div[div == float("inf")] = 0.0
return div
return torch.nan_to_num(div, nan=0.0, posinf=0.0, neginf=0.0)

# %% ../../nbs/losses.pytorch.ipynb 7
def _weighted_mean(losses, weights):
Expand Down Expand Up @@ -825,7 +823,7 @@ def student_scale_decouple(output, loc=None, scale=None, eps: float = 0.1):
if (loc is not None) and (scale is not None):
mean = (mean * scale) + loc
tscale = (tscale + eps) * scale
df = 2.0 + F.softplus(df)
df = 3.0 + F.softplus(df)
return (df, mean, tscale)


Expand Down