rzen 9fd56b6063 Add degradation experiments: one-parameter-at-a-time ablations ...

Controlled study of how the model degrades. Four variants, each = v1 with a
single field replaced (dataclasses.replace), so differences are attributable:
- small: capacity down (2L/64d, ~0.38M params)
- short: training time down (max_iters 20k->2k)
- ctx:   context window down (256->64)
- data:  data quantity down (max_train_tokens->1M; new TrainConfig knob + train.py slice)

- scripts/compare.py: sample same prompt across all trained configs with a
  shared seed, print side by side, write reports/compare.md
- tests/test_configs.py: enforces one-parameter-at-a-time (only intended fields
  differ from v1) + small param count (3 tests). Full suite: 29 passing.

2026-07-12 12:58:48 -04:00

..

test_configs.py

Add degradation experiments: one-parameter-at-a-time ablations

2026-07-12 12:58:48 -04:00

test_dataset.py

Phase 2: token dataset + corpus encoder (code only, not yet run)

2026-07-12 10:38:20 -04:00

test_model.py

Phase 3: decoder-only transformer (~4.3M params)

2026-07-12 10:48:55 -04:00

test_sample.py

Phase 5: autoregressive sampler (temperature + top-k)

2026-07-12 11:34:45 -04:00

test_tokenizer.py

Phase 1: from-scratch byte-level BPE tokenizer

2026-07-12 10:31:16 -04:00

test_train.py

Phase 4: training loop (code only — for user to run)

2026-07-12 10:51:58 -04:00