Algorithms.py

import lightning as L
from lightning.pytorch.utilities import rank_zero_only
import torch
import os
import gc
torch.set_float32_matmul_precision("high")

from SimulateOnEnv import batch_simulate_on_environment
from lightning.pytorch.callbacks import Callback
from transformers import AutoModelForCausalLM, AutoTokenizer, AutoConfig
from typing import Optional
from peft import LoraConfig, TaskType, get_peft_model, PeftModel

from safetensors import safe_open
from safetensors.torch import save_file

def safe_load(path):
    """安全加载权重，处理大小不匹配"""
    result = {}
    with safe_open(path, framework="pt") as f:
        for key in f.keys():
            try:
                result[key] = f.get_tensor(key)
            except Exception as e:
                print(f"Error loading {key}: {str(e)}")
    return result


def set_special_tokens(model, tokenizer):
    if tokenizer.pad_token is None and tokenizer.pad_token_id is None:
        print_rank_0(f"[WARNING] the pad token of the tokenizer is None")
        # We do not resize the vocab embedding, since it ruins the KL value with the ref_model
        tokenizer.pad_token_id = tokenizer.eos_token_id
        tokenizer.pad_token = tokenizer.eos_token
        # tokenizer.pad_token = tokenizer.decode(0)

    model.config.pad_token_id = tokenizer.pad_token_id
    model.config.bos_token_id = tokenizer.bos_token_id
    model.config.eos_token_id = tokenizer.eos_token_id

    return model, tokenizer

def load_model_and_tokenizer(model_name_or_path, actor_checkpoint=None):

    model = AutoModelForCausalLM.from_pretrained(
        model_name_or_path,
        trust_remote_code=True,
        use_cache=False,
        torch_dtype=torch.bfloat16,
        low_cpu_mem_usage=True,
    )
    
    if hasattr(model, "ref_model"):
        del model.ref_model
    
    lora_config = LoraConfig(
        r=8,
        lora_alpha=16,
        target_modules=["q_proj", "v_proj"],
        lora_dropout=0.05,
        bias="none",
        task_type=TaskType.CAUSAL_LM,
    )
    model = get_peft_model(model, lora_config) 
    
    if actor_checkpoint is not None:
        weight_map = {}
        with safe_open(actor_checkpoint, framework="pt") as f:
            for key in f.keys():
                new_key = key.replace("base_model.model.", "")
                weight_map[new_key] = f.get_tensor(key)
        
        # 应用权重
        for name, param in model.named_parameters():
            for key, tensor in weight_map.items():
                if key in name and param.shape == tensor.shape:
                    param.data.copy_(tensor)
                    print(f"加载权重: {name} <- {key}")
                    break
        
    tokenizer = AutoTokenizer.from_pretrained(
        model_name_or_path,
        padding_side="left",  # for batch decode
        truncation_side="left",
        model_max_length=1024,
        trust_remote_code=True,
    )

    model.gradient_checkpointing_enable()
    model, tokenizer = set_special_tokens(model, tokenizer)
    
    return model, tokenizer


class ActorModel(torch.nn.Module):
    def __init__(self, get_device, model_name_or_path, actor_checkpoint=None):
        super().__init__()
        self.get_device = get_device
        self.model, self.tokenizer = load_model_and_tokenizer(model_name_or_path, actor_checkpoint)

    def forward(self, observation, do_sample=True):
        obs_ids = self.tokenizer(
            observation,
            return_tensors="pt",
            padding=True,
            truncation=True,
            max_length=512,
        ).to(self.model.device)
        obs_embeds = self.model.get_input_embeddings()(obs_ids["input_ids"])
        outputs = self.model.generate(
            inputs_embeds=obs_embeds,
            attention_mask=obs_ids["attention_mask"],
            max_new_tokens=32,
            do_sample=do_sample,
            pad_token_id=self.tokenizer.eos_token_id,
        )
        action = self.tokenizer.batch_decode(outputs, skip_special_tokens=True)
        return action

    def behavioral_cloning_loss(self, observation, action, **kwargs):
        logsum_probs = self.get_logsum_prob(
            observation, action
        )  # this line has been refactored and not tested
        loss = -logsum_probs.mean()
        return loss, {"behavioral_cloning/loss": loss.detach()}

    def get_logsum_prob(self, observation, action_from_dataloader, **kwargs):
        action = [a + self.tokenizer.eos_token for a in action_from_dataloader]
        alltext = [obs + a for obs, a in zip(observation, action)]
        generated_probabilities = self.to_tokens_and_logprobs(alltext)
        assert (
            len(generated_probabilities)
            == len(alltext)
            == len(observation)
            == len(action)
        )
        mask = torch.zeros_like(generated_probabilities.detach(), dtype=torch.bool)

        for i, (obs, act, text) in enumerate(zip(observation, action, alltext)):
            assert text == obs + act
            act_ids = self.tokenizer(act, return_tensors="pt", padding=True)
            txt_ids = self.tokenizer(text, return_tensors="pt", padding=True)
            n_token_act = len(
                act_ids["input_ids"][0]
            )  # [0] because the batch is one inside the foor loop
            n_token_txt = len(txt_ids["input_ids"][0])
            mask[i, n_token_txt - n_token_act - 1 : n_token_txt - 1] = (
                True  # the -1 shift is due to the the generated probabilities being shifted
            )

        generated_probabilities = torch.where(mask, generated_probabilities, 1.0)
        log_probs = torch.where(
            mask, torch.log(generated_probabilities), 0.0
        )  # must be separate from the line above for numerical stability (cannot take log(0.0))
        logsum_probs = torch.sum(log_probs, dim=1)
        del act_ids, txt_ids, log_probs, generated_probabilities
        return logsum_probs

    def to_tokens_and_logprobs(self, input_texts):
        input_ids = self.tokenizer(
            input_texts, padding=True, truncation=True, return_tensors="pt"
        ).input_ids.to(self.get_device())
        outputs = self.model(input_ids)
        probs = torch.softmax(outputs.logits, dim=-1)

        # collect the probability of the generated token -- probability at index 0 corresponds to the token at index 1
        probs = probs[:, :-1, :]
        input_ids = input_ids[:, 1:]
        gen_probs = torch.gather(probs, 2, input_ids[:, :, None]).squeeze(-1)
        del outputs, probs
        torch.cuda.empty_cache()
        gc.collect()
        torch.cuda.memory._set_allocator_settings('max_split_size_mb:32')
        return gen_probs


class RobertaCritic(torch.nn.Module):
    def __init__(
        self,
        get_device,
        discount_factor: float,
        tau: float,
        expectile: float,
        from_checkpoint=None,
    ):
        super().__init__()

        self.get_device = get_device
        self.discount_factor = discount_factor
        self.tau = tau
        self.expectile = expectile

        ### Define the Critic
        from ArcherCritic import ArcherDoubleCritic

        self.critic = ArcherDoubleCritic(in_dim=768, out_dim=1)
        self.target_critic = ArcherDoubleCritic(in_dim=768, out_dim=1)
        self.soft_update_target_critic(1)

        if from_checkpoint is not None:
            checkpoint = torch.load(from_checkpoint, map_location=torch.device("cpu"))
            weights = {
                k.removeprefix("critic."): v
                for k, v in checkpoint["state_dict"].items()
                if k.startswith("critic.")
            }
            self.load_state_dict(weights)
            print(
                "I have initialized the critic from the checkpoint: ", from_checkpoint
            )

        ### Miscellaneus Shortcuts
        self.softmax = torch.nn.Softmax(dim=-1)
        self.td_criterion = torch.nn.MSELoss()
        self.expectile_criterion = lambda diff: self.loss_value_diff(
            diff=diff, expectile=self.expectile
        )

    def get_q(self, observation, action, detach_model=False):
        return self.critic.get_q(observation, action, detach_model=detach_model)

    def get_v(self, inputs, detach_model=False):
        return self.critic.get_v(inputs, detach_model=detach_model)

    def get_target_v(self, inputs, detach_model=False):
        return self.target_critic.get_v(inputs, detach_model=detach_model)

    def get_target_q(self, observation, action, detach_model=False):
        return self.target_critic.get_q(observation, action, detach_model=detach_model)

    def get_advantages(self, observation, action):
        q1, q2 = self.get_q(observation, action)
        v1, v2 = self.get_v(observation)
        q = torch.minimum(q1, q2)
        v = torch.minimum(v1, v2)
        advantages = q - v
        return advantages

    def argmax_advantage(self, observation, get_available_actions):
        argmax_actions = []
        for obs in observation:
            available_actions = get_available_actions(obs)
            advantages = torch.as_tensor(
                [self.get_advantages([obs], [action]) for action in available_actions]
            )
            action = available_actions[torch.argmax(advantages)]
            argmax_actions.append(action)
        return argmax_actions

    def soft_update_target_critic(self, tau=None):
        if tau == None:
            tau = self.tau
        for target_param, param in zip(
            self.target_critic.parameters(), self.critic.parameters()
        ):
            target_param.data.copy_(target_param.data * (1.0 - tau) + param.data * tau)

    def iql_loss(self, observation, action, reward, next_observation, done, **kwargs):
        ### Fitting the Q function
        q1, q2 = self.get_q(observation, action, detach_model=False)
        q1 = q1.flatten()
        q2 = q2.flatten()

        reward = torch.Tensor(reward)  # .to(self.agent.device)
        done = torch.Tensor(done)  # .to(self.agent.device)

        with torch.no_grad():
            target_v1, target_v2 = self.get_target_v(next_observation)

            target_v1 = (
                reward
                + torch.logical_not(done) * target_v1.flatten() * self.discount_factor
            )
            target_v2 = (
                reward
                + torch.logical_not(done) * target_v2.flatten() * self.discount_factor
            )

        q1_loss = self.td_criterion(q1, target_v1)
        q2_loss = self.td_criterion(q2, target_v2)

        ### Fitting the value function
        with torch.no_grad():
            target_q1, target_q2 = self.get_target_q(
                observation, action, detach_model=False
            )
        target_q1 = target_q1.flatten()
        target_q2 = target_q2.flatten()

        v1, v2 = self.get_v(observation, detach_model=False)
        v1 = v1.flatten()
        v2 = v2.flatten()

        v1_loss = self.expectile_criterion(diff=target_q1.detach() - v1)
        v2_loss = self.expectile_criterion(diff=target_q2.detach() - v2)

        loss = q1_loss + q2_loss + v1_loss + v2_loss

        ### Log and print what's happening
        log = self.get_log(
            q1=q1,
            q2=q2,
            v1=v1,
            v2=v2,
            q1_loss=q1_loss,
            q2_loss=q2_loss,
            v1_loss=v1_loss,
            v2_loss=v2_loss,
            target_q1=target_q1,
            target_q2=target_q2,
        )
        return loss, log

    def loss_value_diff(self, diff, expectile):
        """Loss function for iql expectile value difference."""
        weight = torch.where(diff > 0, expectile, (1 - expectile))
        return (weight * (diff**2)).mean()

    def get_log(
        self, q1, q2, v1, v2, q1_loss, q2_loss, v1_loss, v2_loss, target_q1, target_q2
    ):
        return {
            "critic/q1.loss": q1_loss.detach(),
            "critic/q2.loss": q2_loss.detach(),
            "critic/v1.loss": v1_loss.detach(),
            "critic/v2.loss": v2_loss.detach(),
            "critic/q1.mean": torch.mean(q1).detach(),
            "critic/q1.min": torch.min(q1).detach(),
            "critic/q1.max": torch.max(q1).detach(),
            "critic/q2.mean": torch.mean(q2).detach(),
            "critic/q2.max": torch.max(q2).detach(),
            "critic/q2.min": torch.min(q2).detach(),
            "critic/v1.mean": torch.mean(v1).detach(),
            "critic/v1.min": torch.min(v1).detach(),
            "critic/v1.max": torch.max(v1).detach(),
            "critic/v2.mean": torch.mean(v2).detach(),
            "critic/v2.max": torch.max(v2).detach(),
            "critic/v2.min": torch.min(v2).detach(),
            "critic/target_q1.mean": torch.mean(target_q1).detach(),
            "critic/target_q1.min": torch.min(target_q1).detach(),
            "critic/target_q1.max": torch.max(target_q1).detach(),
            "critic/target_q2.mean": torch.mean(target_q2).detach(),
            "critic/target_q2.max": torch.max(target_q2).detach(),
            "critic/target_q2.min": torch.min(target_q2).detach(),
        }


class Agent(L.LightningModule):
    def validation_step(self, batch, batch_idx):

        # Perform evaluation on environment with stochastic policy
        return None
        eval_dataset = batch_simulate_on_environment(
            policy=lambda obs: self.forward(obs), env=None
        )
        self.log(
            "eval/avg_return", eval_dataset.mean_trajectory_return(), sync_dist=True
        )
        self.log(
            "eval/std_return", eval_dataset.std_trajectory_return(), sync_dist=True
        )

        # Perform evaluation on environment with deterministic policy
        deterministic_eval_dataset = batch_simulate_on_environment(
            policy=lambda obs: self.forward(obs, do_sample=False),
            env=None,
        )
        self.log(
            "eval/avg_return_deterministic",
            deterministic_eval_dataset.mean_trajectory_return(),
            sync_dist=True,
        )
        self.log(
            "eval/std_return_deterministic",
            deterministic_eval_dataset.std_trajectory_return(),
            sync_dist=True,
        )

        return eval_dataset.mean_trajectory_return()


class BehaviouralCloning(Agent):
    def __init__(self, lr: float):
        super().__init__()  # Initialize LLM base class
        self.save_hyperparameters()

        ### Config
        self.lr = lr

        ### Initialization
        self.agent = ActorModel(get_device=lambda: self.device)

    def forward(self, observation, **kwargs):
        return self.agent.forward(observation, **kwargs)

    def training_step(self, batch, batch_idx):
        loss, log = self.agent.behavioral_cloning_loss(**batch)
        self.log_dict(log, sync_dist=True)
        return loss

    def configure_optimizers(self):
        from torch.optim import Adam
        
        # 收集所有需要优化的参数
        optimizer_params = [
            {"params": self.actor.model.parameters(), "lr": self.actor_lr},
        ]
        
        # 如果需要优化critic，添加其参数
        if self.optimize_critic:
            optimizer_params.append({
                "params": self.critic.critic.parameters(), 
                "lr": self.critic_lr
            })
        
        # 创建单个优化器
        optimizer = Adam(optimizer_params)
        
        return optimizer


class FilteredBehaviouralCloning(BehaviouralCloning):
    def __init__(self, lr: float, filter: float):
        super().__init__(lr)

        self.filter = filter

    def configure_callbacks(self):
        return FilterDataset(filter=self.filter)


class FilterDataset(Callback):
    def __init__(self, filter: float):
        self.filter = filter

    def on_fit_start(self, trainer, algorithm):
        print("*** Filtering Dataset ***")
        dataset = trainer.datamodule.dataset
        print("Statistics of Input Dataset")
        print("Number of Trajectories:", dataset.nTrajectories())
        print("Number of Trajectories:", len(dataset))
        dataset.keep_top_fraction_of_trajectories(fraction=self.filter)
        trainer.datamodule.dataset = dataset
        print("Statistics of Filtered Dataset")
        print("Number of Trajectories:", dataset.nTrajectories())
        print("Number of Trajectories:", len(dataset))


class ActorCritic(Agent):
    def __init__(
        self,
        model_name_or_path: str,
        actor_lr: float,
        critic_lr: float,
        tau: float,
        accumulate_grad_batches: int,
        discount_factor: float,
        critic_expectile: float,
        optimize_critic: bool,
        actor_checkpoint=None,
        critic_checkpoint=None,
        **kwargs
    ):
        super().__init__()  # Initialize LLM base class
        self.example_input_array = (torch.zeros(1, 1, dtype=torch.long),)
        self.save_hyperparameters()
        ### Config
        self.actor_lr = actor_lr
        self.critic_lr = critic_lr
        self.discount_factor = discount_factor
        self.tau = tau

        ### Manual Gradient Accumulation
        self.accumulate_grad_batches = accumulate_grad_batches
        self.automatic_optimization = False

        ### Initialization
        self.actor = ActorModel(
            get_device=lambda: self.device, model_name_or_path=model_name_or_path, actor_checkpoint=actor_checkpoint
        )
        self.critic = RobertaCritic(
            get_device=lambda: self.device,
            discount_factor=discount_factor,
            tau=tau,
            expectile=critic_expectile,
            from_checkpoint=critic_checkpoint,
        )
        
        self.actor_current_backward_step = 0
        self.critic_current_backward_step = 0
        self.critic_warmup_gradient_steps = 0

        self.optimize_actor = lambda: (
            True
            if self.critic_current_backward_step // self.accumulate_grad_batches
            >= self.critic_warmup_gradient_steps
            else False
        )
        self.optimize_critic = lambda: optimize_critic

    def forward(self, observation, **kwargs):
        action = self.actor.forward(observation, **kwargs)
        return action
    
    def training_step(self, batch, batch_idx):
        # if batch_idx == 3:
        #     return
                    
        optimizer = self.optimizers()
        mem = torch.cuda.memory_allocated() / torch.cuda.get_device_properties(0).total_memory
        if mem > 0.8:
            gc.collect()
            torch.cuda.empty_cache()

        if self.optimize_critic():
            # scale losses by 1/N (for N batches of gradient accumulation)
            critic_loss, critic_log = self.critic_loss(batch)
            critic_loss /= self.accumulate_grad_batches
            self.manual_backward(critic_loss)
            self.critic_current_backward_step += 1
            self.log_dict(critic_log, sync_dist=True)

            # accumulate gradients of N batches
            if self.critic_current_backward_step % self.accumulate_grad_batches == 0:
                optimizer.step()
                optimizer.zero_grad()
                self.critic.soft_update_target_critic(self.tau)
            

        if self.optimize_actor():
            # scale losses by 1/N (for N batches of gradient accumulation)
            
            actor_loss, actor_log = self.actor_loss(batch)
            actor_loss /= self.accumulate_grad_batches
            self.manual_backward(actor_loss)
            self.actor_current_backward_step += 1
            self.log_dict(actor_log, sync_dist=True)

            # accumulate gradients of N batches
            if self.actor_current_backward_step % self.accumulate_grad_batches == 0:
                optimizer.step()
                optimizer.zero_grad()

    def get_actor_log(self, loss, advantages, log_prob):
        return {
            "actor/loss": loss.detach(),
            "actor/advantages.mean": advantages.detach().mean(),
            "actor/advantages.max": torch.max(advantages.detach()),
            "actor/advantages.min": torch.min(advantages.detach()),
            "actor/log_prob.mean": torch.mean(log_prob.detach()),
            "actor/log_prob.max": torch.max(log_prob.detach()),
            "actor/log_prob.min": torch.min(log_prob.detach()),
        }

    def configure_optimizers(self):
        from torch.optim import Adam
        
        optimizer_params = []
        
        if hasattr(self, 'actor') and hasattr(self.actor, 'parameters'):
            optimizer_params.append({
                "params": self.actor.parameters(), 
                "lr": self.actor_lr
            })
        
        if self.optimize_critic and hasattr(self, 'critic') and hasattr(self.critic, 'parameters'):
            optimizer_params.append({
                "params": self.critic.parameters(), 
                "lr": self.critic_lr
            })
        
        if not optimizer_params:
            return None
        
        optimizer = Adam(optimizer_params)
        return optimizer


class OfflineArcher(ActorCritic):
    def __init__(
        self,
        model_name_or_path: str,
        inv_temp: float,
        actor_lr: float,
        critic_lr: float,
        tau: float,
        accumulate_grad_batches: int,
        discount_factor: float,
        critic_expectile: float,
        optimize_critic: bool,
        actor_checkpoint: Optional[str] = None,
        critic_checkpoint: Optional[str] = None,
        **kwargs
    ):
        super().__init__(
            model_name_or_path=model_name_or_path,
            actor_lr=actor_lr,
            critic_lr=critic_lr,
            tau=tau,
            accumulate_grad_batches=accumulate_grad_batches,
            discount_factor=discount_factor,
            critic_expectile=critic_expectile,
            optimize_critic=optimize_critic,
            actor_checkpoint=actor_checkpoint,
            critic_checkpoint=critic_checkpoint,
            **kwargs
        )

        self.inv_temp = inv_temp

        self.actor_loss = lambda batch: self.awr_loss(**batch)
        self.critic_loss = lambda batch: self.critic.iql_loss(**batch)

    def awr_loss(self, observation, action, **kwargs):
        log_prob = self.actor.get_logsum_prob(observation, action)
        with torch.no_grad():
            advantages = self.critic.get_advantages(observation, action)

        advantages = advantages.flatten()
        log_prob = log_prob.flatten()
        factor = torch.exp(self.inv_temp * advantages)
        loss_batch = -factor * log_prob
        loss = loss_batch.mean()

        # ### Log and print what's happening
        log = self.get_actor_log(loss=loss, advantages=advantages, log_prob=log_prob)
        log = {
            **log,
            **{
                "actor/factor.mean": factor.detach().mean(),
                "actor/factor.max": torch.max(factor.detach()),
                "actor/factor.min": torch.min(factor.detach()),
            },
        }

        return loss, log
    
    def configure_optimizers(self):
        # 直接调用基类方法
        return super().configure_optimizers()
    
    @rank_zero_only
    def on_save_checkpoint(self, checkpoint):
        """保存 LoRA 适配器权重"""
        super().on_save_checkpoint(checkpoint)
        
        save_dir = self.trainer.default_root_dir
        os.makedirs(save_dir, exist_ok=True)
        
        # 保存 LoRA 适配器
        if hasattr(self.actor.model, "save_pretrained"):
            self.actor.model.save_pretrained(save_dir)
        
        # 保存 tokenizer
        if hasattr(self.actor, "tokenizer"):
            self.actor.tokenizer.save_pretrained(save_dir)
        
        print(f"✅ LoRA adapter saved to: {save_dir}")
        self.merge_and_save_lora(os.path.join(save_dir, "merged_model"))
        
    def merge_and_save_lora(self, save_dir):
        """
        Merge the LoRA adapter weights into the base model and save the merged model and tokenizer.
        """
        # Only proceed if the actor model has the correct method
        try:
            # 确保模型在CPU上且处于eval模式
            original_device = next(self.actor.model.parameters()).device
            self.actor.model.to('cpu')
            self.actor.model.eval()
            
            if hasattr(self.actor.model, "merge_and_unload"):
                # 执行合并
                merged_model = self.actor.model.merge_and_unload()
                
                # 检查合并结果
                from peft import PeftModel
                if isinstance(merged_model, PeftModel):
                    print(">>> [Warning] Still a PeftModel after merge. Using base_model.model...")
                    merged_model = merged_model.base_model.model
                
                # 保存合并后的模型
                merged_model.save_pretrained(os.path.join(save_dir, "merged_model"))
                print(f"✅ Merged model saved to: {os.path.join(save_dir, 'merged_model')}")
            else:
                print("❌ merge_and_unload method not found in actor.model. Cannot merge LoRA weights.")
        except Exception as e:
            print(f"❌ Error merging LoRA weights: {e}")
            import traceback
            traceback.print_exc()
        finally:
            # 恢复原始设备
            self.actor.model.to(original_device)


class IQLKL(ActorCritic):
    def __init__(self, kl_coeff: float, reference_actor_path, **kwargs):
        super().__init__(**kwargs)

        self.kl_coeff = kl_coeff
        self.reference_actor = ActorModel(
            get_device=lambda: self.device, from_checkpoint=reference_actor_path
        )

        self.actor_loss = lambda batch: self.advantage_kl_loss(**batch)
        self.critic_loss = lambda batch: self.critic.iql_loss(**batch)

    def advantage_kl_loss(self, observation, **kwargs):
        reinforce_loss, generated_output = self.reinforce_loss(observation=observation)
        with torch.no_grad():
            reference_log_prob = self.reference_actor.get_logsum_prob(
                observation, generated_output["action"]
            )

        ratio = generated_output["log_prob"] - reference_log_prob
        kl_loss = (ratio.detach() + 1.0) * generated_output["log_prob"]
        loss = (1 - self.kl_coeff) * reinforce_loss + self.kl_coeff * kl_loss
        log = generated_output["log"]
        log = {
            **log,
            "reference_log_prob.mean": reference_log_prob.mean(),
            "reference_log_prob.max": reference_log_prob.max(),
            "reference_log_prob.min": reference_log_prob.min(),
        }
        log = {
            **log,
            "kl_loss.mean": kl_loss.mean(),
            "kl_loss.max": kl_loss.max(),
            "kl_loss.min": kl_loss.min(),
        }
        log = {**log, "actor_loss.mean": loss.mean(), "ratio": ratio.mean()}

        return loss.mean(), log

    def reinforce_loss(self, observation, **kwargs):
        ### Reinforce Loss
        action = self.actor.forward(observation)
        log_prob = self.actor.get_logsum_prob(observation, action)

        with torch.no_grad():
            advantages = self.critic.get_advantages(observation, action)

        loss = -advantages.flatten() * log_prob

        ### Logging
        log = self.get_actor_log(
            loss=torch.mean(loss.detach()), advantages=advantages, log_prob=log_prob
        )
        # self.log_dict(log)
        return loss, {
            "log_prob": log_prob,
            "advantages": advantages,
            "action": action,
            "log": log,
        }