liqiang
b119135836
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s
25 lines
5.7 KiB
XML
25 lines
5.7 KiB
XML
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 40 780 340" width="780" height="340" style="background:#ffffff">
|
|
<defs><marker id="ah" markerWidth="12" markerHeight="8" refX="12" refY="4" orient="auto"><polygon points="0 0, 12 4, 0 8" fill="#333333"/></marker><marker id="ah-light" markerWidth="12" markerHeight="8" refX="12" refY="4" orient="auto"><polygon points="0 0, 12 4, 0 8" fill="#999999"/></marker></defs>
|
|
|
|
<rect x="40" y="65" width="300" height="130" rx="6" fill="#f0f0f0" stroke="#333333" stroke-width="2"/>
|
|
<text x="190" y="88" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="20" fill="#333333" text-anchor="middle" dominant-baseline="central" font-weight="bold">Fase 1: Formateo de SFT</text>
|
|
<text x="190" y="115" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="11" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Objetivo: Salida analizable (JSON/llamada a herramienta)</text>
|
|
<text x="190" y="137" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="13.5" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Datos: Miles de demostraciones de alta calidad</text>
|
|
<text x="190" y="159" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="10" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Condición de parada: Formato estable, capacidad básica lograda</text>
|
|
<text x="190" y="181" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="8" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">⚠ Sobreentrenamiento → El modelo colapsa a la distribución de entrenamiento</text>
|
|
<line x1="342" y1="130" x2="398" y2="130" stroke="#333333" stroke-width="2" marker-end="url(#ah)"/>
|
|
<text x="370" y="115" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="14" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Formato</text>
|
|
<text x="370" y="135" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="14" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Estable</text>
|
|
<rect x="400" y="65" width="340" height="130" rx="6" fill="#d0d0d0" stroke="#333333" stroke-width="2"/>
|
|
<text x="570" y="88" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="19.5" fill="#333333" text-anchor="middle" dominant-baseline="central" font-weight="bold">Fase 2: Estrategia de moldeo de RL</text>
|
|
<text x="570" y="115" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="10" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Objetivo: Maximizar la recompensa de la tarea (precisión/tasa de éxito)</text>
|
|
<text x="570" y="137" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="10.5" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Prerrequisito: Formato de salida estable → Recompensa computable</text>
|
|
<text x="570" y="159" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="9.5" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Avance: Descubrir nuevas estrategias más allá de las demostraciones de SFT</text>
|
|
<text x="570" y="181" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="9.5" fill="#333333" text-anchor="middle" dominant-baseline="central" font-weight="normal">✓ Formato estable + Generalización de estrategia = Listo para despliegue</text>
|
|
<rect x="40" y="220" width="700" height="70" rx="4" fill="#f5f5f5" stroke="#333333" stroke-width="2"/>
|
|
<text x="390" y="238" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="16" fill="#333333" text-anchor="middle" dominant-baseline="central" font-weight="bold">¿Por qué no podemos omitir SFT e ir directamente a RL?</text>
|
|
<text x="390" y="262" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="8" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Salida del modelo base no estructurada → No se puede procesar JSON → La función de recompensa devuelve NaN → Gradientes todos en cero → El entrenamiento falla por completo</text>
|
|
<text x="390" y="310" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="20" fill="#333333" text-anchor="middle" dominant-baseline="central" font-weight="bold">SFT memoriza la distribución → RL generaliza la estrategia</text>
|
|
<text x="390" y="338" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="15.5" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">SFT: máx Σ log P(y|x) (ajustar distribución de entrenamiento) RL: máx E[R(τ)] (optimizar objetivo de la tarea)</text>
|
|
<text x="390" y="362" font-family="Arial, 'Helvetica Neue', Helvetica, 'PingFang SC', 'Microsoft YaHei', sans-serif" font-size="11" fill="#666666" text-anchor="middle" dominant-baseline="central" font-weight="normal">Cuando "no importa cuántas demostraciones se añadan, los nuevos escenarios siguen teniendo un rendimiento bajo" → Punto de inflexión para cambiar a RL</text>
|
|
</svg> |
