giahy2507/speechlm-pretrain-public

Packed speech pretraining corpora One config per corpus, each a set of parquet shards whose rows carry the audio of every segment of one pretraining record, so a corpus can be streamed front to back: from datasets import load_dataset ds = load_dataset("giahy2507/speechlm-pretrain-public", "ami", split="train", streaming=True) Rows: id, source, seconds, num_segments, meta (JSON), and segments, a list of {audio: bytes, text, seconds, speaker, meta}. audio is an encoded clip… See the full description on the dataset page: https://huggingface.co/datasets/giahy2507/speechlm-pretrain-public.

<iframe src="https://silytics.si/embed/item/b4293c2f-1d68-4045-a2d2-f4601f98bc4e" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.