ceselder/easynla-dsv4-warmstart-opus5
EasyNLA warm-start for DeepSeek-V4-Flash-0731 — Opus-5 explanations NLA (Natural Language Autoencoder) warm-start data: DeepSeek-V4-Flash-0731 layer-28 activations (last token of finefineweb prefixes; docs/positions identical to asher577/easynla-warmstart-data) paired with gold explanations written by claude-opus-5 (same instruction prompt as the original Sonnet-4.6 set; thinking disabled, max_tokens 400; 742,661 requests, 62 fallbacks). Measured effect vs the Sonnet-4.6… See the full description on the dataset page: https://huggingface.co/datasets/ceselder/easynla-dsv4-warmstart-opus5.
Re-split: unified train/test (99/1, doc-level crc32) replaces the AV/AR-disjoint layout
Upload ar_sft_val.parquet.nla_meta.yaml with huggingface_hub
Upload ar_sft_shuf.parquet.nla_meta.yaml with huggingface_hub
Upload av_sft_val.parquet.nla_meta.yaml with huggingface_hub
Upload av_sft_shuf.parquet.nla_meta.yaml with huggingface_hub
Upload corpus/ar_val.parquet with huggingface_hub
Upload corpus/av_val.parquet with huggingface_hub
Upload corpus/ar_train.parquet with huggingface_hub
Upload corpus/av_train.parquet with huggingface_hub
Upload ar_sft_val.parquet with huggingface_hub
Upload ar_sft_shuf.parquet with huggingface_hub
Upload av_sft_val.parquet with huggingface_hub
Upload av_sft_shuf.parquet with huggingface_hub
Upload README.md with huggingface_hub
initial commit