Upload model, telemetry and hardware context
Run context assets
Model Artifact
{% set model_asset = uploads.model %} {% if model_asset %} {% if model_asset.description %}{{ model_asset.description }}
{% endif %} {% else %}Upload a model checkpoint to anchor telemetry comparisons.
{% endif %}Hardware Inventory
{% set hardware_asset = uploads.hardware %} {% if hardware_asset %} {% if hardware_asset.summary %}-
{% if hardware_asset.summary.cpu %}
- CPU {{ hardware_asset.summary.cpu }} {% endif %} {% if hardware_asset.summary.memory %}
- Memory {{ hardware_asset.summary.memory }} {% endif %} {% if hardware_asset.summary.gpus %}
- GPUs
-
{% for gpu in hardware_asset.summary.gpus %}
- {{ gpu.name or "GPU" }}{% if gpu.count %} ×{{ gpu.count }}{% endif %}{% if gpu.memory %} — {{ gpu.memory }}{% endif %} {% endfor %}
{% endif %}
{{ hardware_asset.preview }}
{% endif %}
{% else %}
Upload a hardware snapshot (JSON or text) to capture the cluster shape.
{% endif %}Telemetry & Usage
Baseline Telemetry
{% set baseline = uploads.telemetry.baseline %} {% if baseline %} {% if baseline.metrics %}-
{% for metric in baseline.metrics %}
- {{ metric.label }} {% if metric.value is number %} {{ "%.3f"|format(metric.value) }} {% else %} {{ metric.value }} {% endif %} {% endfor %}
Bottleneck: {{ baseline.bottleneck.name }} {% if baseline.bottleneck.latency_ms is not none %} ({{ "%.3f"|format(baseline.bottleneck.latency_ms) }} ms) {% endif %}
{% endif %}Layers profiled: {{ baseline.event_count }}
{% if baseline.preview %}{{ baseline.preview }}
{% endif %}
{% else %}
Upload baseline telemetry JSON to anchor comparisons.
{% endif %}Optimized Telemetry
{% set optimized = uploads.telemetry.optimized %} {% if optimized %} {% if optimized.metrics %}-
{% for metric in optimized.metrics %}
- {{ metric.label }} {% if metric.value is number %} {{ "%.3f"|format(metric.value) }} {% else %} {{ metric.value }} {% endif %} {% endfor %}
Bottleneck: {{ optimized.bottleneck.name }} {% if optimized.bottleneck.latency_ms is not none %} ({{ "%.3f"|format(optimized.bottleneck.latency_ms) }} ms) {% endif %}
{% endif %}Layers profiled: {{ optimized.event_count }}
{% if optimized.preview %}{{ optimized.preview }}
{% endif %}
{% else %}
Upload optimized telemetry JSON to track gains.
{% endif %}Usage Event
{% set usage = uploads.usage %} {% if usage %} {% if usage.highlights %}-
{% for highlight in usage.highlights %}
- {{ highlight.label }} {% if highlight.value is number %} {{ "%.3f"|format(highlight.value) }} {% else %} {{ highlight.value }} {% endif %} {% endfor %}
{{ usage.preview }}
{% endif %}
{% else %}
Optional: upload a billing or usage event to capture savings.
{% endif %}Performance snapshot
{% if summary.metrics %}| Metric | Baseline | Optimized | Trend | Delta | Delta % |
|---|---|---|---|---|---|
| {{ metric.label }} | {{ metric.baseline if metric.baseline is not none else "—" }} | {{ metric.optimized if metric.optimized is not none else "—" }} | {% if max_value > 0 %} {% else %} No data {% endif %} | {% if metric.delta is not none %} {{ "%.3f"|format(metric.delta) }} {% else %} — {% endif %} | {% if metric.delta_pct is not none %} {{ "%.2f"|format(metric.delta_pct) }}% {% else %} — {% endif %} |
Upload baseline and optimized telemetry to populate this view.
{% endif %}Logs & Notes
{% if uploads.logs %}{{ log.description }}
{% endif %} {% if log.preview %}{{ log.preview }}
{% endif %}
Attach runtime logs or analyst notes to keep the optimization trail together.
{% endif %}Layer-wise statistics
{% if layer_stats %}| Layer | Baseline latency (ms) | Optimized latency (ms) | Metadata |
|---|---|---|---|
| {{ layer.name }} | {{ layer.baseline_latency_ms if layer.baseline_latency_ms is not none else "—" }} | {{ layer.optimized_latency_ms if layer.optimized_latency_ms is not none else "—" }} |
{% if layer.metadata %}
{{ layer.metadata | tojson(indent=2) }}
{% else %}
—
{% endif %}
|
No layer telemetry detected. Upload telemetry JSON with an `events` array.
{% endif %}Kernel artifacts
{% if artifacts %}{% if artifact.media_type %} {% endif %} {% if artifact.size %} {% endif %}
Upload optimized IR or generated kernels to stage downloads.
{% endif %}Benchmarks comparison
{% if summary.metrics %}| Metric | Baseline | Optimized | Trend | Delta | Delta % |
|---|---|---|---|---|---|
| {{ metric.label }} | {{ metric.baseline if metric.baseline is not none else "—" }} | {{ metric.optimized if metric.optimized is not none else "—" }} | {% if max_value > 0 %} {% else %} No data {% endif %} | {% if metric.delta is not none %} {{ "%.3f"|format(metric.delta) }} {% else %} — {% endif %} | {% if metric.delta_pct is not none %} {{ "%.2f"|format(metric.delta_pct) }}% {% else %} — {% endif %} |
Upload usage or benchmark data to compare performance.
{% endif %}Export SDK pack
{% if summary.metrics %}Generate a bundle containing telemetry, usage events, kernel artifacts and a performance summary.
{% else %}Provide telemetry to enable the export bundle.
{% endif %}