mfu-kit

name	mfu-kit
description	Use MFUKit for global model FLOPs utilization logging, including microbatch FLOPs accumulation, device peak resolution, distributed reduction, and perf/mfu log payload construction.

MFU Kit

Goal

Use MFUKit to log optimizer-step model FLOPs utilization:

• accumulate_microbatch(...) records local logical model FLOPs;
• build_log(...) resolves peak TFLOPs, reduces global FLOPs and step time, and returns {"perf/mfu": ...};
• recipe/model code still owns architecture-specific calculate_model_flops.

Required Inputs

• where the recipe computes real batch metadata;
• a model method or explicit value for local microbatch FLOPs;
• step timing at optimizer-step boundary;
• device type, effective precision, and optional configured device peak.

Workflow

1. Initialize self.mfu_kit = MFUKit().
2. Add a recipe model patch that attaches calculate_model_flops(...) when the model does not provide it.
3. In forward_step(), call accumulate_microbatch(model=..., batch_size=..., seq_len=..., **metadata).
4. In post-step logging, call build_log(device_type=..., precision=..., step_time_seconds=...) and merge the returned payload into logs.

Validation

Soft Validation

• FLOPs use actual batch metadata, not only config defaults;
• gradient accumulation microbatches are included exactly once;
• step time is the synchronized optimizer-step time;
• effective precision and device peak are explicit or resolved by the kit;
• missing peak data results in no MFU log rather than a guessed value.

Output

• State where FLOPs are computed, accumulated, and logged.
• State device/precision/peak source and validation status.

Read On Demand

• skills/training/model-flops-utilization/SKILL.md: feature-oriented workflow and formula details.