robometer/rewardeval_ui
4
1o
2 �Ti;3 � @ sd d dl mZmZmZmZ d dlZd dlZd dlm Z m4Z5 d dl m Z d dl
mZ G dd� de �ZdS )� )�Dict�List�Any�OptionalN)�ProgressSample�6Trajectory)�RFMBaseSampler)� rank_0_printc sj e Zd ZdZ ddee f� fdd�
Zdeee e7f fdd�Z d e de
fd8d �Zd d
� Zdd� Z� ZS )�ProgressDefaultSamplerztDataset that generates progress samples by iterating through each trajectory in the dataset, used in policy ranking.N�max_trajectoriesc s^ t � jdi |�� || _tdt| j�� d�| jd� | �� | _tdt| j�� d�| jd� d S )Nz(ProgressDefaultSampler initialized with �
trajectories��verbosez9Generated z sample indices� ) �super�__init__r r �len�robot_trajectoriesr � _generate_all_sample_indices�sample_indices)�selfr �kwargs�� __class__r �B/scr/aliang80/reward_fm/rfm/data/samplers/eval/progress_default.pyr
s �10 z ProgressDefaultSampler.__init__�returnc C s� | j }| jdur| jt| j �k r| j�| j | j�}tdt|�� d�| jd� g }|D ]}|�|| j| d | j| d d�� q*|S )z%Generate all possible sample indices.Nz(Generating progress default samples for r r
�frames�id)�traj_idx�11video_pathr ) r r r �
_local_random�sampler r �append�dataset)r �trajectories_to_processr �ir r r r s �(z3ProgressDefaultSampler._generate_all_sample_indices�sample_idx_infoc C sV |d }|d }| j | }|d |d |d |d |d�}| j||d�}t|d �}|S ) z8Generate a single progress sample from trajectory index.r r �
quality_label� data_source�taskr )r' r( r) r r )�traj�metadata)�12trajectoryN)r# �_get_traj_from_datar )r r& r r r* r+ r, r! r r r � _generate_sample_from_indices, s 13��14z4ProgressDefaultSampler._generate_sample_from_indicesc C s15 t | j�S �N)r r )r r r r �__len__E s 16z ProgressDefaultSampler.__len__c C s | � | j| �S r/ )r. r )r �idxr r r � __getitem__H s z"ProgressDefaultSampler.__getitem__r/ )�__name__�17__module__� __qualname__�__doc__r �intr r r �strr r �dictr r. r0 r2 �
__classcell__r r r r r18 19 s ��r20 )�typingr r r r �numpy�np�torch�rfm.data.dataset_typesr r �rfm.data.samplers.baser �rfm.utils.distributedr r21 r r r r �<module> s