From 4a0a7a26d831c8de2e10f19446d5eed975eb4638 Mon Sep 17 00:00:00 2001 From: tatp-yf Date: Fri, 10 Apr 2026 16:36:03 +0800 Subject: [PATCH] =?UTF-8?q?refactor:=20unify=20observation=20dimension=20n?= =?UTF-8?q?aming=20n=5Fobs=20=E2=86=92=20obs=5Fdim?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fixes #190 --- src/unilab/algos/torch/common/networks.py | 10 +++++----- src/unilab/algos/torch/fast_td3/learner.py | 12 ++++++------ src/unilab/utils/algo_utils.py | 2 +- 3 files changed, 12 insertions(+), 12 deletions(-) diff --git a/src/unilab/algos/torch/common/networks.py b/src/unilab/algos/torch/common/networks.py index 2f16227be..fccab20b9 100644 --- a/src/unilab/algos/torch/common/networks.py +++ b/src/unilab/algos/torch/common/networks.py @@ -18,7 +18,7 @@ class DistributionalQNetwork(nn.Module): def __init__( self, - n_obs: int, + obs_dim: int, n_act: int, num_atoms: int, v_min: float, @@ -28,7 +28,7 @@ def __init__( ): super().__init__() self.net = nn.Sequential( - nn.Linear(n_obs + n_act, hidden_dim, device=device), + nn.Linear(obs_dim + n_act, hidden_dim, device=device), nn.ReLU(), nn.Linear(hidden_dim, hidden_dim // 2, device=device), nn.ReLU(), @@ -96,7 +96,7 @@ class Critic(nn.Module): def __init__( self, - n_obs: int, + obs_dim: int, n_act: int, num_atoms: int, v_min: float, @@ -106,7 +106,7 @@ def __init__( ): super().__init__() self.qnet1 = DistributionalQNetwork( - n_obs=n_obs, + obs_dim=obs_dim, n_act=n_act, num_atoms=num_atoms, v_min=v_min, @@ -115,7 +115,7 @@ def __init__( device=device, ) self.qnet2 = DistributionalQNetwork( - n_obs=n_obs, + obs_dim=obs_dim, n_act=n_act, num_atoms=num_atoms, v_min=v_min, diff --git a/src/unilab/algos/torch/fast_td3/learner.py b/src/unilab/algos/torch/fast_td3/learner.py index ec6cef5ef..19ea20fe1 100644 --- a/src/unilab/algos/torch/fast_td3/learner.py +++ b/src/unilab/algos/torch/fast_td3/learner.py @@ -46,7 +46,7 @@ class TD3Actor(nn.Module): def __init__( self, - n_obs: int, + obs_dim: int, n_act: int, num_envs: int, init_scale: float, @@ -58,7 +58,7 @@ def __init__( super().__init__() self.n_act = n_act self.net = nn.Sequential( - nn.Linear(n_obs, hidden_dim, device=device), + nn.Linear(obs_dim, hidden_dim, device=device), nn.ReLU(), nn.Linear(hidden_dim, hidden_dim // 2, device=device), nn.ReLU(), @@ -175,7 +175,7 @@ def __init__( # Build actor self.actor = TD3Actor( - n_obs=obs_dim, + obs_dim=obs_dim, n_act=action_dim, num_envs=num_envs, init_scale=init_scale, @@ -185,7 +185,7 @@ def __init__( device=torch_device, ) self.actor_target = TD3Actor( - n_obs=obs_dim, + obs_dim=obs_dim, n_act=action_dim, num_envs=num_envs, init_scale=init_scale, @@ -198,7 +198,7 @@ def __init__( # Build critic self.qnet = Critic( - n_obs=obs_dim, + obs_dim=obs_dim, n_act=action_dim, num_atoms=num_atoms, v_min=v_min, @@ -207,7 +207,7 @@ def __init__( device=torch_device, ) self.qnet_target = Critic( - n_obs=obs_dim, + obs_dim=obs_dim, n_act=action_dim, num_atoms=num_atoms, v_min=v_min, diff --git a/src/unilab/utils/algo_utils.py b/src/unilab/utils/algo_utils.py index a518f92e4..679af2493 100644 --- a/src/unilab/utils/algo_utils.py +++ b/src/unilab/utils/algo_utils.py @@ -41,7 +41,7 @@ def build_actor( from unilab.algos.torch.fast_td3.learner import TD3Actor return TD3Actor( - n_obs=obs_dim, + obs_dim=obs_dim, n_act=action_dim, num_envs=num_envs, hidden_dim=actor_hidden_dim,