Convert Euler angles (roll, pitch, yaw) to quaternion. The order of Euler angles is yaw-pitch-roll (Z-Y-X axis rotation order).

order textured, mesh {obj} then mesh.ply

Convienience function to find the Median Absolute Deviation from a vector of deviations. Note that it isn't a strict median( just a quick approximatio

! For n != 0, returns k, such that pow10 := 10^(k-1) <= n < 10^k. For n == 0, returns 1 and sets pow10 := 1. */

! For n != 0, returns k, such that pow10 := 10^(k-1) <= n < 10^k. For n == 0, returns 1 and sets pow10 := 1. */

Get all the logged statistics for the given dataset.

Dimensionality of the policy's action space.

使用 GPT-4 模型识别图像中的地标

! For a normalized diyfp w = f * 2^e, this function returns a (normalized) cached power-of-ten c = f_c * 2^e_c, such that the exponent of the product

! For a normalized diyfp w = f * 2^e, this function returns a (normalized) cached power-of-ten c = f_c * 2^e_c, such that the exponent of the product

Either computes the statistics of a dataset or loads them from a cache file if this function has been called before with the same `hash_depen

Return (width, height) for a given img file content - no external dependencies except the os and struct builtin modules

Returns a list of indices so that each slice of `global_batch_size` consecutive indices corresponds to elements of the same modality

Generates dataset kwargs for a given dataset mix from the Open X-Embodiment dataset. The returned kwargs (per-dataset configs) and weights ca

Return the scale factor needed to go from current_res to target_res. The value is calculated based on the idea that we don't want to crop the inpu

Return the base name of a textured mesh obtained with RealityCapture Args: source_folder (str): path to dataset (bundle file, textured an

Initialize RLDS Dataset (wraps TFDS), ActionTokenizer, and initialize transform/collation functions.

! Generates V = buffer * 10^decimal_exponent, such that M- <= V <= M+. M- and M+ must be normalized and share the same exponent -60 <= e <= -32. */

! Generates V = buffer * 10^decimal_exponent, such that M- <= V <= M+. M- and M+ must be normalized and share the same exponent -60 <= e <= -32. */