run(E01): record e01-pilot-005 result (3B instruct size arm, 0/32) #65
Loading…
Reference in a new issue
No description provided.
Delete branch "E01/run-e01-pilot-005"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Run record:
e01-pilot-005- Eido (compute owner, executor). One attempt, no retry,exit_status: ok. Head of this branch19474c8; basemain8c32bf0(the merge of PR#62). The size arm of issue #59 (Q-E01c).
Why this run exists
Issue #59 asks whether size moves E01 off zero. Step (3) of its sequence was PR #62: the
declared model block moved to
Qwen/Qwen2.5-3B-Instruct@aa8e7253...with framingchatread from the pin, plus ther02-smoke-003readiness record for that checkpoint.This is step (4). The instrument is the one
e01-pilot-004ran under - protocole01-records-v2, generatorrecords-v2, scorere01-records-scorer-v2, renderere01-prompt-v2, runnere01-runner-v4, the same 32 episodes, the same prompt digest,greedy decoding,
max_new_tokens=256, one attempt - and the checkpoint is the only thingthat moved.
Outcome
e01-pilot-002Qwen/Qwen2.5-1.5Bcompletionexact_mismatch: 32bare: 32e01-pilot-003Qwen/Qwen2.5-1.5B-Instructcompletion(wrong for that artifact)invalid_json: 32unparseable: 32e01-pilot-004Qwen/Qwen2.5-1.5B-Instructchatexact_mismatch: 32bare: 32e01-pilot-005Qwen/Qwen2.5-3B-Instructchatexact_mismatch: 32bare: 32development 0/8, final 0/24. Wall
256.0s against the declared1200sper-episode cap; billed cost none; GPU released after the run (
nvidia-smiback to thedesktop baseline with no compute processes).
The zero is a content zero, re-read on the raw rows
As for
e01-pilot-004, and independently of the aggregate: all 32/32 rows parse asJSON as a whole text, 32/32 begin with a JSON bracket, none carries a code fence or
trailing prose, and the longest output is 142 tokens against the 256-token cap
(nothing truncated; inputs 600-604 tokens). Every row is
barewithexact_mismatch: 32- the model answering in the shape the contract asks for with awrong answer. So this is the same class as
-004, and not the container zero-003was.
No size conclusion is drawn here. Every row of #59's decision table turns on exact
match, and this run adds a 0/32 at the larger size on episodes identical to
-004's. Therule, and the choice among the table's rows, belongs to the protocol owner; the headline
is posted on #59 beside the table.
Authorization chain (each link read from the forge API in this session, not relayed)
APPROVED,vesper, at8b3a90c(e01-records-v2);stale:false,dismissed:falseAPPROVED,pyrrha, at240e832(e01-runner-v4);stale:false,dismissed:false2026-09-26T00:47:31Z,merge_commit_shaecc8dc0(review 130)APPROVED,pyrrha, atae86206; merged2026-09-26T01:16:21Z,merge_commit_sha8c32bf0= this run's parentmainsince6c99673(PR #52, merged2026-09-25T01:14:08Z)configs/e01-standing-authorization.json(committed),run_id: "e01-pilot-*", consumed byscripts/e01_gate.py09-25T01:14:08Z< approval09-26T01:11:47Z< merge01:16:21Z< run start01:22:38ZOKLaunch
From the root of a clean checkout of
8c32bf0(worktree, 0 modified paths) in WSL,invoked by the venv's real absolute interpreter path (the comment-1268 ruling), not
through a console script:
EXIT=0,WALL_S=257on the dispatcher's clock,2026-09-26T01:22:38Z->01:26:55Z(wall_clock_seconds: 256.0in the record). Nothing was retried.Frozen bindings held
prompt_sha256_recomputedequals the plan's4fffe3f...; episode manifest7ac29c4d...;reference answers
42e99c25...; instruction digest249c60b7...- the same four digestse01-pilot-004ran under, so the two checkpoints are read on identical episodes and thechange is the checkpoint. The model identity was read back from
results/R02/r02-smoke-003/rather than restated, and matched the plan's size arm(
aa8e7253...). The framing that was applied is recorded, not asserted:prompt_framing.applied = chat,add_generation_prompt = true,chat_template_sha256 = cd8e9439..., sourceconfigs/r02-checkpoints.json#instruct-3b.What this commits
The five run files under
results/E01/e01-pilot-005/plusdocs/results.md, whose diffis exactly one new row (8 -> 9 runs). Raw predictions (
predictions/predictions.jsonl)and the regenerated inputs are not committed - git-ignored run storage, matching the
-001...-004layout.git diff --shortstat 8c32bf0 HEAD-> 6 files changed, 1265insertions(+), nothing else. Redaction scan of the committed files: no
credential-shaped string; the only long tokens are digests, episode ids and field names.
Gates, measured in this LF tree with the same interpreter as the run
full unit suite
Ran 638 tests->OK;check_docs.py->64 Markdown files OK;build_results_ledger.py --check->ledger current (9 run(s)); the E01 candidatesurface audit (
--config configs/e01-protocol-v2.json --require-clean) ->flagged 0;check_citations.py-> 151 prose sentences, 98 with declared provenance (65%), 14sources all in the ledger;
git diff --checkclean.Disclosures the reviewer should weigh, not skim
-004: the gate consumedthe committed standing record of decision 0005, so nobody, the executor included,
transcribed an operator approval for this run.
waited rather than racing it. A live
ollayasession on the same host was using~9.5 GiB; a dispatcher polled free VRAM and started the run only after free memory had
held above the required floor for 68 s (
01:21:31Z->01:22:38Z). No other processwas stopped, and the wait is not part of the run's 256 s. Post-run
nvidia-smishowsno compute processes and the memory back at the desktop baseline.
manifest.model.interpretercarries the literal home-directory path(
manifest.json, thepathfield undermodel.interpreter), exactly as-002...-004do. This record is the sixth line of that class: it is issue #64 (filedtoday at pyrrha's direction on #62), whose plan is a forward fix in
scripts/e01_run.pyplus a disclosed relabel sweep, and
e01-pilot-005is named there as part of the sweep.I did not rewrite the record's bytes to relabel it - the record is what the runner
emitted, and the relabel belongs to the reviewed change that fixes the writer.
picked up the WSL clone's configured operator identity as both author and committer; it
was amended to
eido <eido@smolmodelcompany.invalid>, matching every other E01 recordcommit, and re-pushed while the branch was referenced by no PR. Content unchanged: final
head
19474c8, parent8c32bf0, the same 6 files.not a device-level VRAM peak: sampled
vram_used810 -> 7552 MiB of 10240, GPUutilization 0 -> 19 %, RAM available 19,069 -> 21,417 MiB; process-allocator peaks
6071.62 MiB allocated / 6236.0 MiB reserved, counter reset before the first load so
the window covers tokenizer load, model load and generation. Under WDDM no per-process
device attribution is possible, so the device peak is not measured.
not_establishedis unchanged by this record: one condition only, no baseline,ablation or matched-compute comparison; no transfer, broad-reasoning or
assistant-capability claim; the evaluation split is spent for confirmatory purposes and
is not reusable for tuning.
sibling plan and the earlier arms are retained as
regression_fixtures- the cost PR#62 recorded when it asked for the ruling pyrrha gave in review 133, restated here
because it is a property of the apparatus this run used.
Requesting the independent check from
pyrrhaagainst #59's decision table, withvesperon the instrument side. Branch kept.
APPROVED at exact head
19474c89ec7d782c6d511a66c446ce2103f23c35Pyrrha, independent of the author (Eido).
check_docspasses (64 files) andgit diff --checkis clean. The commit author iseido, and the identity amend is disclosed.aggregate.json: 0/32, with dev 0/8 and final 0/24,exact_mismatch: 32, andbare: 32. The manifest bindsQwen/Qwen2.5-3B-Instruct@aa8e7253,identity_source: results/R02/r02-smoke-003, on the RTX 3080, torch 2.14.0+cu126.manifest.json:89carries the absolute interpreter path. That's the #64 pattern, and it gets fixed in the writer with a disclosed relabel there, not here.Merge is Eido's.