You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
ColossalAI/applications/Chat/coati/experience_maker/naive.py

72 lines
2.8 KiB

import torch
import torch.nn.functional as F
from coati.models.base import Actor, Critic, RewardModel
from coati.models.generation import generate
from coati.models.utils import calc_action_log_probs, compute_reward
from transformers import PreTrainedTokenizer
from .base import Experience, ExperienceMaker
class NaiveExperienceMaker(ExperienceMaker):
"""
Naive experience maker.
"""
def __init__(
self,
actor: Actor,
critic: Critic,
reward_model: RewardModel,
initial_model: Actor,
tokenizer: PreTrainedTokenizer,
kl_coef: float = 0.1,
) -> None:
super().__init__(actor, critic, reward_model, initial_model)
self.tokenizer = tokenizer
self.kl_coef = kl_coef
@torch.no_grad()
def make_experience(self, input_ids: torch.Tensor, **generate_kwargs) -> Experience:
self.actor.eval()
self.critic.eval()
self.initial_model.eval()
self.reward_model.eval()
# generate sequences
sequences = generate(self.actor, input_ids, self.tokenizer, **generate_kwargs)
# calculate auxiliary tensors
attention_mask = None
pad_token_id = self.tokenizer.pad_token_id
if pad_token_id is not None:
attention_mask = sequences.not_equal(pad_token_id).to(dtype=torch.long, device=sequences.device)
input_len = input_ids.size(1)
eos_token_id = self.tokenizer.eos_token_id
if eos_token_id is None:
action_mask = torch.ones_like(sequences, dtype=torch.bool)
else:
# left padding may be applied, only mask action
action_mask = (sequences[:, input_len:] == eos_token_id).cumsum(dim=-1) == 0
action_mask = F.pad(action_mask, (1 + input_len, -1), value=True) # include eos token and input
action_mask[:, :input_len] = False
action_mask = action_mask[:, 1:]
action_mask = action_mask[:, -(sequences.size(1) - input_len) :]
num_actions = action_mask.size(1)
actor_output = self.actor(sequences, attention_mask)["logits"]
action_log_probs = calc_action_log_probs(actor_output, sequences, num_actions)
base_model_output = self.initial_model(sequences, attention_mask)["logits"]
base_action_log_probs = calc_action_log_probs(base_model_output, sequences, num_actions)
value = self.critic(sequences, attention_mask)
r = self.reward_model(sequences, attention_mask)
reward = compute_reward(r, self.kl_coef, action_log_probs, base_action_log_probs, action_mask=action_mask)
advantage = reward - value
# TODO(ver217): maybe normalize adv
if advantage.ndim == 1:
advantage = advantage.unsqueeze(-1)
return Experience(sequences, action_log_probs, value, reward, advantage, attention_mask, action_mask)