torchaudio
torchaudio copied to clipboard
R interface to torchaudio
Hi ! i try to run this dataset but can not [https://skeydan.github.io/Deep-Learning-and-Scientific-Computing-with-R-torch/audio_classification.html](url) ``` library(torch) library(torchaudio) library(luz) ds
- [x] tuneR_loader - [x] av_loader - [x] transform_to_tensor - [x] transform_to_tensor.Wave - [ ] transform_to_tensor.numeric - [ ] transform_to_tensor.matrix
models
``` r Sys.time() [1] "2023-03-25 14:21:28 GMT" library(magrittr) url % httr::content() %>% rvest::html_element("table") %>% rvest::html_table() %>% dplyr::mutate( model = X1 %>% stringr::str_c("- [ ] ", .) ) %>% dplyr::pull(model) %>%...
i just wanted to point out that there is something off in the NFCC calculations in torchaudio that is not present in pytorchaudio and therefore seems to not be carried...
``` r Sys.time() #> [1] "2020-10-29 01:25:49 -03" library(magrittr) url % httr::content() %>% xml2::xml_find_all('//span[text()="def"]/following-sibling::span[1]') %>% xml2::xml_text() %>% stringr::str_c("- [ ] ", ., "\n") %>% t() %>% cat() ``` - [x]...
``` r Sys.time() [1] "2020-11-02 01:12:45 -03" library(magrittr) url % httr::content() %>% xml2::xml_find_all('//*[@id="datasets"]/ul/li') %>% xml2::xml_text() %>% stringr::str_c("- [ ] ", ., "\n") %>% t() %>% cat() ``` - [x] CMUARCTIC...
``` r Sys.time() [1] "2020-11-02 04:01:57 -03" library(magrittr) url % httr::content() %>% xml2::xml_find_all('//span[text()="def"]/following-sibling::span[1]') %>% xml2::xml_text() %>% stringr::str_c("- [ ] ", ., "\n") %>% t() %>% cat() ``` - [ ]...
``` r Sys.time() [1] "2020-10-30 18:44:17 -03" library(magrittr) url % httr::content() %>% xml2::xml_find_all('//span[text()="class"]/following-sibling::span[1]') %>% xml2::xml_text() %>% stringr::str_c("- [ ] ", ., "\n") %>% t() %>% cat() ``` - [x] Spectrogram...
I'm using utils::untar() into dataset classes to extract donwloaded files (). This function do not allow control over overwrite option.