cyy/go2_rl_gym
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

v0.1.6; add rem-cts

Browse Source
This commit is contained in:
wty-yy
2026-01-07 23:01:52 +08:00
parent 858314fcfa
commit 41295f7e72
10 changed files with 419 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
rsl_rl/rsl_rl/algorithms/__init__.py
View File

@@ -33,4 +33,5 @@ from .cts import CTS
from .moe_cts import MoECTS
from .mcp_cts import MCPCTS
from .ac_moe_cts import ACMoECTS
from .dual_moe_cts import DualMoECTS
from .dual_moe_cts import DualMoECTS
from .rem_cts import REMCTS

234
rsl_rl/rsl_rl/algorithms/rem_cts.py Normal file
View File

@@ -0,0 +1,234 @@
# SPDX-FileCopyrightText: Copyright (c) 2021 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: BSD-3-Clause
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
#
# 1. Redistributions of source code must retain the above copyright notice, this
# list of conditions and the following disclaimer.
#
# 2. Redistributions in binary form must reproduce the above copyright notice,
# this list of conditions and the following disclaimer in the documentation
# and/or other materials provided with the distribution.
#
# 3. Neither the name of the copyright holder nor the names of its
# contributors may be used to endorse or promote products derived from
# this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
# SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
# CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
# OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
# Copyright (c) 2021 ETH Zurich, Nikita Rudin
import torch
import torch.nn as nn
import torch.optim as optim
import itertools
from rsl_rl.modules import ActorCriticMoECTS
from rsl_rl.storage import RolloutStorageCTS
from rsl_rl.algorithms.cts import CTS
class REMCTS(CTS):
model: ActorCriticMoECTS
def __init__(self,
model,
num_envs,
history_length,
num_learning_epochs=1,
num_mini_batches=1,
clip_param=0.2,
gamma=0.998,
lam=0.95,
value_loss_coef=1.0,
entropy_coef=0.0,
load_balance_coef=0.01,
learning_rate=1e-3,
student_encoder_learning_rate=1e-3,
max_grad_norm=1.0,
use_clipped_value_loss=True,
schedule="fixed",
desired_kl=0.01,
teacher_env_ratio=0.75,
device='cpu',
):
self.device = device
self.desired_kl = desired_kl
self.schedule = schedule
self.learning_rate = learning_rate
self.history_length = history_length
# CTS components
self.model = model
self.model.to(self.device)
self.storage = None # initialized later
params1 = [
{"params": self.model.teacher_encoder.parameters()},
{"params": self.model.critic.parameters()},
{"params": self.model.actor.parameters()},
{"params": self.model.std}
]
self.optimizer1 = optim.Adam(params1, lr=learning_rate)
self.optimizer2 = optim.Adam(self.model.student_moe_encoder.parameters(), lr=student_encoder_learning_rate)
self.transition = RolloutStorageCTS.Transition()
# CTS parameters
self.clip_param = clip_param
self.num_learning_epochs = num_learning_epochs
self.num_mini_batches = num_mini_batches
self.value_loss_coef = value_loss_coef
self.entropy_coef = entropy_coef
self.load_balance_coef = load_balance_coef
self.gamma = gamma
self.lam = lam
self.max_grad_norm = max_grad_norm
self.use_clipped_value_loss = use_clipped_value_loss
self.teacher_num_envs = max(int(num_envs * teacher_env_ratio), 1)
self.student_num_envs = num_envs - self.teacher_num_envs
student_env_ratio = 1 - teacher_env_ratio
self.teacher_env_idxs = torch.tensor([i for i in range(num_envs) if i % int(1/student_env_ratio) != 0], device=self.device)
self.student_env_idxs = torch.tensor([i for i in range(num_envs) if i % int(1/student_env_ratio) == 0], device=self.device)
assert len(self.teacher_env_idxs) == self.teacher_num_envs, f"{len(self.teacher_env_idxs)=} != {self.teacher_num_envs=}"
assert len(self.student_env_idxs) == self.student_num_envs, f"{len(self.student_env_idxs)=} != {self.student_num_envs=}"
def update(self):
mean_value_loss = 0
mean_surrogate_loss = 0
mean_entropy_loss = 0
mean_latent_loss = 0
mean_load_balance_loss = 0
assert not self.model.is_recurrent
data = list(self.storage.mini_batch_generator(self.num_mini_batches, self.num_learning_epochs))
teacher_samples = self.teacher_num_envs * self.storage.num_transitions_per_env // self.num_mini_batches
student_samples = self.student_num_envs * self.storage.num_transitions_per_env // self.num_mini_batches
for sample in data:
(
obs_batch, privileged_obs_batch, actions_batch, history_batch,
target_values_batch, advantages_batch, returns_batch,
old_actions_log_prob_batch, old_mu_batch, old_sigma_batch,
hid_states_batch, masks_batch
) = sample
def get_results(start, end, is_teacher):
self.model.act(obs_batch[start:end], privileged_obs_batch[start:end], history_batch[start:end], is_teacher)
actions_log_prob = self.model.get_actions_log_prob(actions_batch[start:end])
value = self.model.evaluate(privileged_obs_batch[start:end], history_batch[start:end], is_teacher)
mu = self.model.action_mean
sigma = self.model.action_std
entropy = self.model.entropy
return actions_log_prob, value, mu, sigma, entropy
teacher_results = get_results(0, teacher_samples, True)
student_results = get_results(teacher_samples, teacher_samples + student_samples, False)
results = []
for x1, x2 in zip(teacher_results, student_results):
results.append(torch.cat([x1, x2], dim=0))
actions_log_prob_batch = results[0]
value_batch = results[1]
mu_batch = results[2]
sigma_batch = results[3]
entropy_batch = results[4]
# KL
if self.desired_kl != None and self.schedule == 'adaptive':
with torch.inference_mode():
kl = torch.sum(
torch.log(
sigma_batch / old_sigma_batch + 1.e-5) + (
torch.square(old_sigma_batch) +
torch.square(old_mu_batch - mu_batch)
) / (2.0 * torch.square(sigma_batch)) - 0.5, axis=-1)
kl_mean = torch.mean(kl)
if kl_mean > self.desired_kl * 2.0:
self.learning_rate = max(1e-5, self.learning_rate / 1.5)
elif kl_mean < self.desired_kl / 2.0 and kl_mean > 0.0:
self.learning_rate = min(1e-2, self.learning_rate * 1.5)
for param_group in self.optimizer1.param_groups:
param_group['lr'] = self.learning_rate
# Surrogate loss
ratio = torch.exp(actions_log_prob_batch - torch.squeeze(old_actions_log_prob_batch))
surrogate = -torch.squeeze(advantages_batch) * ratio
surrogate_clipped = -torch.squeeze(advantages_batch) * torch.clamp(ratio, 1.0 - self.clip_param,
1.0 + self.clip_param)
surrogate_losses = torch.max(surrogate, surrogate_clipped)
teacher_surrogate_loss = surrogate_losses[:teacher_samples].mean()
student_surrogate_loss = surrogate_losses[teacher_samples:].mean()
surrogate_loss = teacher_surrogate_loss + student_surrogate_loss
# surrogate_loss = teacher_surrogate_loss
# Value function loss
if self.use_clipped_value_loss:
value_clipped = target_values_batch + (value_batch - target_values_batch).clamp(-self.clip_param,
self.clip_param)
value_losses = (value_batch - returns_batch).pow(2)
value_losses_clipped = (value_clipped - returns_batch).pow(2)
value_loss = torch.max(value_losses, value_losses_clipped).mean()
else:
value_loss = (returns_batch - value_batch).pow(2).mean()
# teacher_value_loss = value_losses[:teacher_samples].mean()
# student_value_loss = value_losses[teacher_samples:].mean()
# value_loss = teacher_value_loss # + student_value_loss
loss = surrogate_loss + self.value_loss_coef * value_loss - self.entropy_coef * entropy_batch.mean()
# Gradient step
self.optimizer1.zero_grad()
loss.backward()
params_to_clip = itertools.chain.from_iterable(g['params'] for g in self.optimizer1.param_groups)
nn.utils.clip_grad_norm_(params_to_clip, self.max_grad_norm)
self.optimizer1.step()
mean_value_loss += value_loss.item()
mean_surrogate_loss += surrogate_loss.item()
mean_entropy_loss += entropy_batch.mean().item()
for sample in data:
(
obs_batch, privileged_obs_batch, actions_batch, history_batch,
target_values_batch, advantages_batch, returns_batch,
old_actions_log_prob_batch, old_mu_batch, old_sigma_batch,
hid_states_batch, masks_batch
) = sample
# Student encoder update
student_latent, gating_weights = self.model.student_moe_encoder(history_batch[teacher_samples:])
with torch.no_grad():
teacher_latent = self.model.teacher_encoder(privileged_obs_batch[teacher_samples:])
latent_loss = (teacher_latent - student_latent).pow(2).mean()
# Load balance loss
mean_usage = torch.mean(gating_weights, dim=0)
target_usage = torch.full_like(mean_usage, 1.0 / gating_weights.shape[1])
load_balance_loss = torch.mean((mean_usage - target_usage).pow(2))
# load_balance_loss = torch.sum(mean_usage.pow(2)) * gating_weights.shape[1] # Switch Transformer style
student_loss = latent_loss + self.load_balance_coef * load_balance_loss
self.optimizer2.zero_grad()
student_loss.backward()
nn.utils.clip_grad_norm_(self.model.student_moe_encoder.parameters(), self.max_grad_norm)
self.optimizer2.step()
mean_latent_loss += latent_loss.item()
mean_load_balance_loss += load_balance_loss.item()
num_updates = self.num_learning_epochs * self.num_mini_batches
mean_value_loss /= num_updates
mean_surrogate_loss /= num_updates
mean_entropy_loss /= num_updates
mean_latent_loss /= num_updates
mean_load_balance_loss /= num_updates
self.storage.clear()
return mean_value_loss, mean_surrogate_loss, mean_entropy_loss, mean_latent_loss, mean_load_balance_loss