PR 6498 eval status display SVG screenshots

eval-status-complete-clean.svg

eval-status-narrow.svg

eval-status-wide.svg

<svg xmlns="http://www.w3.org/2000/svg" width="1400" height="520" viewBox="0 0 1400 520">
<style>

  .bg { fill: #fafaf8; }
  .chrome { fill: #eeeeec; stroke: #d2d2d2; }
  .txt { font: 18px Menlo, Consolas, monospace; fill: #1f1f1f; }
  .small { font: 16px Menlo, Consolas, monospace; fill: #666; }
  .cyan { fill: #00b2c2; }
  .green { fill: #00a546; }
  .yellow { fill: #b28c00; }
  .muted { fill: #777; }
  .barbg { fill: #dadad6; }
  .bar { fill: #2a2a2a; }
  .okbox { fill: #e8f8ee; stroke: #aadcb9; }

</style>
<rect class="bg" x="0" y="0" width="1400" height="100%"/>
<rect class="chrome" x="0" y="0" width="1400" height="58"/>
<circle cx="24" cy="29" r="7" fill="#ff5f56"/>
<circle cx="48" cy="29" r="7" fill="#ffbd2e"/>
<circle cx="72" cy="29" r="7" fill="#27c93f"/>
<text class="small" x="100" y="35">uv run benchmarking/evals_v3/execute.py --config gdp_pdf.yaml --verbose</text>
<text class="txt cyan" x="24" y="110">[2026-07-08 12:49:16] DEBUG</text>
<text class="txt" x="320" y="110">Eval run mode debug=True upload_to_db=False private=True generation_only=False</text>
<text class="txt cyan" x="24" y="144">[2026-07-08 12:49:16] DEBUG</text>
<text class="txt" x="320" y="144">Batch-enabled host models: openai_gpt-5-4-mini-non-reasoning max_concurrency=1 timeout=600</text>
<text class="txt cyan" x="24" y="204">⠋</text>
<text class="txt cyan" x="58" y="204">gdp_pdf_liteparse_text</text>
<text class="txt muted" x="320" y="204">/</text>
<text class="txt green" x="348" y="204">openai_gpt-5-4-mini-non-reasoning (x1)</text>
<text class="txt muted" x="840" y="204">gen</text>
<rect class="barbg" x="895" y="188" width="385" height="18" rx="3"/><rect class="bar" x="895" y="188" width="284" height="18" rx="3"/>
<text class="txt" x="1296" y="204">74%</text>
<text class="txt cyan" x="24" y="246">⠙</text>
<text class="txt cyan" x="58" y="246">gdp_pdf_liteparse_text</text>
<text class="txt muted" x="320" y="246">/</text>
<text class="txt green" x="348" y="246">openai_gpt-5-4-mini-non-reasoning -&gt; judge_model</text>
<text class="txt muted" x="980" y="246">grade</text>
<rect class="barbg" x="1055" y="230" width="225" height="18" rx="3"/><rect class="bar" x="1055" y="230" width="72" height="18" rx="3"/>
<text class="txt" x="1296" y="246">32%</text>
<line x1="20" y1="308" x2="1380" y2="308" stroke="#e2e2de"/>
<text class="small" x="24" y="354">Wide terminal: full layout uses a flexible bar column and expands to the full available width.</text>
</svg>