> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tuneplane.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Benchmarks

> The score matrix across runs and benchmarks, at /benchmarks.

![TunePlane benchmarks page](https://mintcdn.com/starforge/GatXR2rI5-_Vm4_H/images/console/benchmarks.png?fit=max\&auto=format\&n=GatXR2rI5-_Vm4_H\&q=85\&s=4c566bdcc2e13ced6378a9d975671356)

Benchmarks board. Empty until you run tp bench.

Every score `tp bench` has produced, on one matrix at `/benchmarks`. Comparable across runs,
methods and base models.

## What you see

| Region | What it shows |
| - | - |
| **Rows** | Models and runs |
| **Columns** | Benchmarks: GSM8K, MMLU, C-Eval and the rest |
| **A cell** | The score and the sample count it was measured over |
| **Green** | The best score in that column |
| **Filters** | Base model, method, runner, time range |

The sample count matters. A score measured with `--limit 20` is not comparable with a full one,
and the cell is where you see that.

## What you can do

| Action | Effect |
| - | - |
| Click a score | Opens the evaluation job: its logs and the runner's parameters |
| Filter by base model | The only honest way to read "did post-training help" |
| Run a benchmark | `tp bench run` — see [running a benchmark](https://docs.tuneplane.ai/en/guides/benchmarks.md) |

Benchmark the base model before you train. Without that row the board has no reference frame and
every improvement is a number with nothing to compare against.

## Where the numbers come from

`tp bench` submits an evaluation job that runs lm-eval, evalscope or the RTL harness inside the
platform's evaluation image. Scores are ingested idempotently: re-reporting the same run and
benchmark updates the row rather than adding a second one.

Every runner obeys the same score contract, so the same benchmark is comparable whichever one ran
it. Each metric also carries a declared direction, so the board knows which way is better without
guessing.

Externally produced scores enter the same matrix, labelled as such. See
[reporting from an external harness](https://docs.tuneplane.ai/en/guides/benchmarks.md).

## Next

[Running a benchmark](https://docs.tuneplane.ai/en/guides/benchmarks.md) · [Rubrics](https://docs.tuneplane.ai/en/console/rubrics.md) ·
[Review](https://docs.tuneplane.ai/en/console/review.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.