torchtyping => jaxtyping (#1838)

* Run `ruff --fix`

* Add jaxtyping migration script

* Run migration scripts + ruff + black

* Merge remote-tracking branch 'origin/main' into brent/jaxtyping

* black

* Fix lint

* fix lint

* Cleanup and more specific types

* Fixed annotation errors

* Fix more errors

* Fix annotation

* Delete _torchtyping_to_jaxtyping.py

---------

Co-authored-by: Matt Tancik <mattt4001@gmail.com>
This commit is contained in:
Brent YiandMatt Tancik authored and GitHub committed 2023-05-16 15:57:47 -07:00
1 parent 9b6290ea43
commit 6ba2a18b2b
65 files changed
+597 -531

No files matched your search

+10 -6
View File
@@ -187,11 +187,12 @@ class NerfactoModel(Model):
self.density_fns.extend([network.density_fn for network in self.proposal_networks])
# Samplers
update_schedule = lambda step: np.clip(
np.interp(step, [0, self.config.proposal_warmup], [0, self.config.proposal_update_every]),
1,
self.config.proposal_update_every,
)
def update_schedule(step):
return np.clip(
np.interp(step, [0, self.config.proposal_warmup], [0, self.config.proposal_update_every]),
1,
self.config.proposal_update_every,
)
# Change proposal network initial sampler if uniform
initial_sampler = None # None is for piecewise as default (see ProposalNetworkSampler)
@@ -244,7 +245,10 @@ class NerfactoModel(Model):
def set_anneal(step):
# https://arxiv.org/pdf/2111.12077.pdf eq. 18
train_frac = np.clip(step / N, 0, 1)
bias = lambda x, b: (b * x) / ((b - 1) * x + 1)
def bias(x, b):
return b * x / ((b - 1) * x + 1)
anneal = bias(train_frac, self.config.proposal_weights_anneal_slope)
self.proposal_sampler.set_anneal(anneal)
+11 -6
View File
@@ -140,11 +140,13 @@ class NerfplayerNerfactoModel(NerfactoModel):
self.density_fns.extend([network.density_fn for network in self.proposal_networks])
# Samplers
update_schedule = lambda step: np.clip(
np.interp(step, [0, self.config.proposal_warmup], [0, self.config.proposal_update_every]),
1,
self.config.proposal_update_every,
)
def update_schedule(step):
return np.clip(
np.interp(step, [0, self.config.proposal_warmup], [0, self.config.proposal_update_every]),
1,
self.config.proposal_update_every,
)
self.proposal_sampler = ProposalNetworkSampler(
num_nerf_samples_per_ray=self.config.num_nerf_samples_per_ray,
num_proposal_samples_per_ray=self.config.num_proposal_samples_per_ray,
@@ -246,7 +248,10 @@ class NerfplayerNerfactoModel(NerfactoModel):
if "depth_image" in batch.keys() and self.config.depth_weight > 0:
mask = (batch["depth_image"] != 0).view([-1])
loss_dict["depth_loss"] = 0
l = lambda x: self.config.depth_weight * (x - batch["depth_image"][mask]).pow(2).mean()
def l(x):
return self.config.depth_weight * (x - batch["depth_image"][mask]).pow(2).mean()
loss_dict["depth_loss"] = l(outputs["depth"][mask])
for i in range(self.config.num_proposal_iterations):
loss_dict["depth_loss"] += l(outputs[f"prop_depth_{i}"][mask])
+6 -2
View File
@@ -119,7 +119,8 @@ class NeuSFactoModel(NeuSModel):
self.density_fns.extend([network.density_fn for network in self.proposal_networks])
# update proposal network every iterations
update_schedule = lambda step: -1
def update_schedule(_):
return -1
initial_sampler = UniformSampler(single_jitter=self.config.use_single_jitter)
self.proposal_sampler = ProposalNetworkSampler(
@@ -149,7 +150,10 @@ class NeuSFactoModel(NeuSModel):
def set_anneal(step: int):
# https://arxiv.org/pdf/2111.12077.pdf eq. 18
train_frac = np.clip(step / N, 0, 1)
bias = lambda x, b: (b * x) / ((b - 1) * x + 1)
def bias(x, b):
return b * x / ((b - 1) * x + 1)
anneal = bias(train_frac, self.config.proposal_weights_anneal_slope)
self.proposal_sampler.set_anneal(anneal)
+4 -1
View File
@@ -160,7 +160,10 @@ class SemanticNerfWModel(Model):
def set_anneal(step):
# https://arxiv.org/pdf/2111.12077.pdf eq. 18
train_frac = np.clip(step / N, 0, 1)
bias = lambda x, b: (b * x) / ((b - 1) * x + 1)
def bias(x, b):
return b * x / ((b - 1) * x + 1)
anneal = bias(train_frac, self.config.proposal_weights_anneal_slope)
self.proposal_sampler.set_anneal(anneal)