evidence(ship-007): live apr trace --save-tensor smoke on canonical 7B teacher#1419
Merged
Merged
Conversation
… 7B teacher End-to-end live smoke of `apr trace --save-tensor` against the canonical Qwen2.5-Coder-7B-Instruct-Q4K teacher on RTX 4090 (lambda-labs). Outcome: APRT byte format produced verbatim for `embedding,lm_head` stages. - `layer-0/embedding.bin`: 100,364 bytes = 12-byte header + (7×3584)×4 - `lm_head.bin`: 608,268 bytes = 12-byte header + 152,064×4 xxd-confirmed APRT magic + correct layer/dim header fields. Forward pass succeeded — 28 layer activations + 152,064 logits — same as `apr run`. Discharges three FALSIFY-APR-TRACE-SAVE-* gates from PARTIAL_ALGORITHM_LEVEL to FUNCTIONAL_DISCHARGED level (009/010/011). Contract v1.4.0 bump follows once #1418 v1.3.0 paperwork lands. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
End-to-end live smoke of
apr trace --save-tensoron the canonical Qwen2.5-Coder-7B-Instruct-Q4K teacher (RTX 4090, lambda-labs). APRT byte format produced verbatim for bothembeddingandlm_headstages.Result
xxd confirms APRT magic + correct layer/dim header fields:
embedding.bin: layer=0, dim=25088 (=7×3584 hidden_dim) ✓lm_head.bin: layer=0xFFFFFFFF (WHOLE_MODEL_LAYER), dim=152064 ✓Why
Empirically discharges three FALSIFY-APR-TRACE-SAVE-* gates from PARTIAL_ALGORITHM_LEVEL → FUNCTIONAL_DISCHARGED:
Contract v1.4.0 bump promoting these gates follows once #1418 v1.3.0 paperwork lands. This evidence file is the source of truth for that bump.
Test plan
🤖 Generated with Claude Code