argmaxinc/earnings22-kws-golden

--- dataset_info: features: - name: audio dtype: audio: sampling_rate: 16000 - name: file_id dtype: string - name: transcript list: string - name: text dtype: string - name: norm_text dtype: string - name: dictionary list: string - name: keywords list: string - name: audio_start dtype: float32 - name: audio_end dtype: float32 - name: duration dtype: float32 - name: is_flagged dtype: bool splits: - name: test num_bytes: 371409619.0 num_examples: 772 - name: validation num_bytes: 49058004.0 num_examples: 102 download_size: 404218999 dataset_size: 420467623.0 configs: - config_name: default data_files: - split: test path: data/test-* - split: validation path: data/validation-* --- `earnings22-keywords` Annotation of this dataset is still in progress. Argmax will publish a conference paper on the annotation process later in 2025. The license for this dataset is the same as that of the [original dataset](https://arxiv.org/pdf/2203.15591).