| license: mit | |
| task_categories: | |
| - automatic-speech-recognition | |
| language: | |
| - km | |
| # Khmer ASR Phase 1 Cache | |
| This repository contains cached training artifacts used for Khmer ASR backbone evaluation. | |
| ## Source Datasets | |
| - seanghay/khmer_grkpp_speech | |
| - seanghay/km-speech-corpus | |
| - KrorngAI/fleurs_openslr42_mpwt | |
| ## Contents | |
| - combined_dataset.pt | |
| ## Usage | |
| ```python | |
| import torch | |
| data = torch.load("combined_dataset.pt") |