{% extends "base.html" %} {% block title %}Run History - lm-eval-ledger{% endblock %} {% block content %}
{{ r.source_yaml or r.config_yaml }}
| model | task (few-shot) | acc | n | no answer | size | time | tok/s | |
|---|---|---|---|---|---|---|---|---|
| {{ b.model_tag }} | {{ b.task }}{% if b.fewshot_k is not none %}({{ b.fewshot_k }}){% endif %} | {% if b.verified_accuracy is not none %} {{ "%.3f"|format(b.verified_accuracy) }}✓ {% else %}{{ "%.3f"|format(b.accuracy) if b.accuracy is not none else "…" }}{% endif %} | {% set pending = b.accuracy is none and not b.error %}{{ b.total_examples if b.total_examples is not none else ("…" if pending else "-") }} | {{ b.no_answer_count if b.no_answer_count is not none else ("…" if pending else "-") }} | {% set nb = bench_bytes.get(b.benchmark_id) %}{{ nb|fmtbytes if nb is not none else ("…" if pending else "-") }} | {% if b.error %}ERROR{% elif b.duration_seconds is not none %}{{ b.duration_seconds|fmtdur }}{% elif pending and elapsed.get(b.benchmark_id) is not none %}{{ elapsed[b.benchmark_id]|fmtdur }} …{% else %}{{ "…" if pending else "-" }}{% endif %} | {{ "%.0f"|format(b.gen_tokens / b.gen_seconds) if b.gen_tokens and b.gen_seconds else ("…" if pending else "-") }} |
no benchmarks recorded
{% endif %}The ledger is empty.
{% endfor %} {% endblock %}