Non-contextualized pplx-embed-v1-4B · 512-token chunks · 2560-dim INT8
Chunk-level embeddings of every deduplicated English FineWiki
article, produced with Perplexity's pplx-embed-v1
family. One row per text chunk with its native INT8 embedding and a bit-packed
binary embedding.
Model: perplexity-ai/pplx-embed-v1-4B
Contextualization: NO — each chunk encoded independently
Chunk size: 512 tokens (paragraph-greedy packing)
Embedding… See the full description on the dataset page:
https://huggingface.co/datasets/geopti/finewiki-pplx-embed-base-4B-512tok.