Execution attempt

Sign in with GitHub
← Experiment E5 · How much of the change from an untouched model to the same model after an APT4 transplant and continued pretraining is attributable to the continued pretraining rather than to the tokenizer replacement, when the original-tokenizer model receives identical continued pretraining?

Execution attempt · A35 · planned

Builds the arm-B model: APT4 transplanted into Qwen2.5-1.5B with Fast Vocabulary Transfer initialisation.

Succeeded · Registered by @stw2 via agent. Reported and received times are kept apart; no computation or result is verified.

Pinned source and configuration

https://github.com/stw2/tokenizer-science-tax @ 4d64dedcbe57761c48ea37d5ce80cb6c86198236

Reference checked 2026-09-14 13:10 UTC. Later commits, branches or plan changes do not retarget this attempt.

Command
python scripts/03_transplant_fvt.py
Working directory
experiments/E05-control-arm
Configuration paths
None
Parameters
Each APT4 piece's embedding is the mean of the Qwen rows of its re-encoding; <s> and </s> take the <|endoftext|> row; <unk> and 256 byte pieces take the mean embedding; tied embeddings; vocabulary 32,000; eos 2 and generation pad 2. The digest is that of the scoring machine's copy of the directory, which also holds a special_tokens_map.json the script does not write.
Environment
unrecorded; PyTorch, CUDA 13.1, bf16; one CUDA GPU, 16 GB
Output directory
Not recorded

Inputs

Materials the registrant named when registering this attempt, by content identity; obtainability is derived from their location reports. Nothing is fetched or verified.

Delivered events

  1. Registered

    #1

    Builds the arm-B model: APT4 transplanted into Qwen2.5-1.5B with Fast Vocabulary Transfer initialisation.

    reported · received · @stw2 via agent · posted to the Thread

  2. Started

    #2

    Started.

    reported · received · @stw2 via agent · attempt only

  3. Succeeded

    #3

    Completed with exit code 0.

    Exit code 0.

    • qwen2.5-1.5b-apt4-fvt model directory · Not redistributed: ask the Room owner, or rebuild with scripts/03_transplant_fvt.py at the pinned revisions · restricted · sha256 fa48359351d3… · 2722731769 bytes · M132
    • fvt_stats.json · https://github.com/stw2/tokenizer-science-tax/blob/9a57aa40f4e774136f51b6730c33266d59094a45/experiments/E05-control-arm/results/fvt_stats.json · public · sha256 ae87326684be… · 119 bytes · M133

    reported · received · @stw2 via agent · posted to the Thread

Report an event

The registrant’s capture tool normally delivers start and outcome events. Reporting here is the same author report with the browser as the reported time; it does not observe the process.

This attempt has a delivered outcome. A new execution is a new attempt.