Add NanoBEIR-en results for sbintuitions/sarashina-embedding-v2-1b

Summary

• Adds the 13 NanoBEIR-en compressed result files to the existing Sarashina result directory.
• This supplements the 538 standard-task files merged in PR #23, bringing the model to the restored 551-task --all scope.
• Only hakari-results/sbintuitions__sarashina-embedding-v2-1b/hakari-bench__NanoBEIR-en/*.json.xz is included.

Reproducibility

CUDA_VISIBLE_DEVICES=1 PYTHONPATH="$PWD" PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True tmp/venvs/sarashina-st402-tf4-fa2/bin/python -m hakari_bench.cli evaluate dense --model sbintuitions/sarashina-embedding-v2-1b --model-revision 1f3408afaa7b617e3445d891310a9c26dd0c68a5 --dataset hakari-bench/NanoBEIR-en --dtype bf16 --flash-attn2 --device cuda:0 --batch-size 16 --query-prompt "task: 質問を与えるので、その質問に答えるのに役立つ関連文書を検索してください。\nquery: " --document-prompt "text: " --results-dir output/sarashina-st402-tf4-fa2-nanobeir-en --show-progress

Verification

• 13/13 NanoBEIR-en task files validated.
• Each file records the pinned model revision and the five expected dense embedding variants.
• The target path did not exist on the result repository main branch before this PR.