|
import torch |
|
from diffusers import DiffusionPipeline |
|
import tqdm |
|
|
|
from diffusers.models.unet_1d import UNet1DModel |
|
from diffusers.utils.dummy_pt_objects import DDPMScheduler |
|
|
|
|
|
class ValueGuidedDiffuserPipeline(DiffusionPipeline): |
|
def __init__(self, value_function: UNet1DModel, unet: UNet1DModel, scheduler: DDPMScheduler, env): |
|
super().__init__() |
|
self.value_function = value_function |
|
self.unet = unet |
|
self.scheduler = scheduler |
|
self.env = env |
|
self.data = env.get_dataset() |
|
self.means = dict((key, val.mean(axis=0)) for key, val in self.data.items()) |
|
self.stds = dict((key, val.std(axis=0)) for key, val in self.data.items()) |
|
self.device = self.unet.device |
|
self.state_dim = env.observation_space.shape[0] |
|
self.action_dim = env.action_space.shape[0] |
|
|
|
def normalize(self, x_in, key): |
|
return (x_in - self.means[key]) / self.stds[key] |
|
|
|
def de_normalize(self, x_in, key): |
|
return x_in * self.stds[key] + self.means[key] |
|
|
|
def to_torch(self, x_in): |
|
|
|
if type(x_in) is dict: |
|
return {k: self.to_torch(v) for k, v in x_in.items()} |
|
elif torch.is_tensor(x_in): |
|
return x_in.to(self.device) |
|
return torch.tensor(x_in, device=self.device) |
|
|
|
def reset_x0(self, x_in, cond, act_dim): |
|
for key, val in cond.items(): |
|
x_in[:, key, act_dim:] = val.clone() |
|
return x_in |
|
|
|
def run_diffusion(self, x, conditions, n_guide_steps, scale): |
|
batch_size = x.shape[0] |
|
y = None |
|
for i in tqdm.tqdm(self.scheduler.timesteps): |
|
|
|
timesteps = torch.full((batch_size,), i, device=self.device, dtype=torch.long) |
|
|
|
for _ in range(n_guide_steps): |
|
with torch.enable_grad(): |
|
x.requires_grad_() |
|
y = self.value_function(x, timesteps).sample |
|
grad = torch.autograd.grad([y.sum()], [x])[0] |
|
|
|
posterior_variance = self.scheduler._get_variance(i) |
|
model_std = torch.exp(0.5 * posterior_variance) |
|
grad = model_std * grad |
|
grad[timesteps < 2] = 0 |
|
x = x.detach() |
|
x = x + scale * grad |
|
x = self.reset_x0(x, conditions, self.action_dim) |
|
|
|
prev_x = self.unet(x.permute(0, 2, 1), timesteps).sample.permute(0, 2, 1) |
|
x = self.scheduler.step(prev_x, i, x, predict_epsilon=False)["prev_sample"] |
|
|
|
|
|
x = self.reset_x0(x, conditions, self.action_dim) |
|
x = self.to_torch(x, device=self.device) |
|
|
|
return x, y |
|
|
|
def __call__(self, obs, batch_size=64, planning_horizon=20, n_guide_steps=2, scale=0.1): |
|
obs = self.normalize(obs, "observations") |
|
obs = obs[None].repeat(batch_size, axis=0) |
|
conditions = {0: self.to_torch(obs)} |
|
shape = (batch_size, planning_horizon, self.state_dim + self.action_dim) |
|
x1 = torch.randn(shape, device=self.device) |
|
x = self.reset_x0(x1, conditions, self.action_dim) |
|
x = self.to_torch(x) |
|
x, y = self.run_diffusion(x, conditions, n_guide_steps, scale) |
|
sorted_idx = y.argsort(0, descending=True).squeeze() |
|
sorted_values = x[sorted_idx] |
|
actions = sorted_values[:, :, : self.action_dim] |
|
actions = actions.detach().cpu().numpy() |
|
denorm_actions = self.de_normalize(actions, key="actions") |
|
|
|
denorm_actions = denorm_actions[0, 0] |
|
return denorm_actions |
|
|