verdigris-d2d8b9d2 / README.md
iamPi's picture
Document the cleaning pass.
861c814 verified
|
Raw
History Blame Contribute Delete
2.38 kB
metadata
license: other
task_categories:
  - text-generation
tags:
  - bittensor
  - sn120
  - affine
  - sft
size_categories:
  - 10K<n<100K

verdigris-d2d8b9d2

Cleaned SFT corpus for Bittensor SN120 (Affine), Reason v4 (weight_version_key=7). 17,655 rows, TRL prompt-completion format.

Format is exactly ` THOUGHT: {z}

{y}` on every row.

How it was cleaned

Derived from iamPi/tessera-77d11909. Prompts and actions are byte-identical to the source; only the thought channel changed, and 483 rows (2.66%) were removed.

Repaired in place (40.08% of rows had at least one defect)

defect rows
stray <think> / </think> / <thinking> / </thinking> in the thought 7,124
corrupted THOUGHT label (THO, THOTH:, THOFTWARE:, THOTHought:, THOURLY Thought:, THOproperly …) 218
duplicated leading THOUGHT: / THOUGHT 188
a second THOUGHT label mid-thought 72
in-thought ```bash fence retagged to ```sh 114

The retag matters: the validator's BASH_RE matches only ```bash, and split_rollout takes the last match as the action. A thought that quotes a bash fence could therefore be mistaken for the command whenever a model's real fence came out malformed.

Removed (483 rows)

reason rows
thought shorter than the 80-char crown floor 444
completion longer than the 1,792-token rollout budget 34
harness output (<returncode> / <output>) leaked into the thought 5

Those last five were collapsed multi-turn transcripts: the "thought" contained the result of a command that had not been run yet, which would train the model to hallucinate observations.

Two-channel thoughts — latent analysis followed by a visible summary — were kept. That is the contract's own shape: split_rollout joins the latent and visible halves on purpose, and 57 of the 61 such rows carry genuinely distinct text in each half.

Guarantees

Every row was checked after rewriting: it round-trips through the validator's own split_rollout, is byte-identical to the force_text() rendering, has no think tag / THOUGHT label / bash fence / harness tag in its thought, clears the 80-char floor, and fits the rollout budget.

Variants

A cue-appended build with an identical row set exists for A/B testing.