Week 02 — Track 1 — 40% target
This is the complete evaluation and reproduction package for enrollment
25M2111. The selected checkpoint is the Round-18 tail-Fisher block-adaptive
model, chosen as a transparent post-holdout robustness override.
| Item | Value |
|---|---|
| Base model | Qwen/Qwen3.5-4B |
| Artifact | week02_40_tail_fisher_block64.pt |
| Exact size | 3,359,426,040 bytes |
| 40% ceiling margin | 573,960 bytes |
| SHA256 | fce48abc65e13f86308f6886417c640fb72ea6b48389818b75574221f5717dd0 |
| Week-2 matched checkpoint | 0.918 accuracy on 560 problems |
| Sealed 2026 holdout | 0.603 accuracy on 63 problems |
| GitHub | https://github.com/safffrron/CS6013/tree/main/25M2111/Week02/Track1_40/Submission01 |
| Hugging Face | https://huggingface.co/safffrron/25M2111-Week02-Track1-40-Submission01 |
Evaluate the uploaded checkpoint
Run from this Submission01 directory:
python -m venv .venv
source .venv/bin/activate
python -m pip install -e '.[eval]'
hf download safffrron/25M2111-Week02-Track1-40-Submission01 \
--local-dir hf_checkpoint
sha256sum hf_checkpoint/week02_40_tail_fisher_block64.pt
python dequantize_to_bf16.py \
Qwen/Qwen3.5-4B hf_checkpoint restored_hf
python inference/smoke.py restored_hf
./evaluation/run_public_eval.sh restored_hf 0
The digest must match the table. Evaluation defaults to
max_new_tokens=65536 and repetition penalty 1.20; recorded artifact
evidence used 32,768 tokens.
Required conversion API
code.py provides:
convert_from_hf_checkpoint(model_name, output_path, sparsity=None)convert_to_hf_checkpoint(model_name, checkpoint_path, output_path)
The optional sparsity argument is ignored because this is mixed-precision
quantization, not pruning. checkpoint_path may be the download directory or
the artifact file. Missing files are fetched from the submission repository.
Method, selection, and reproduction
The short-reasoning source is partitioned into 64-row blocks. A nested successive-refinement allocator preserves the Round-16 widths and spends the remaining budget by upgrading 212 answer-tail-Fisher-ranked blocks (36,012,032 weights) from 6 to 8 bits. The physical stream contains dense mixed-width integer codes, FP16 group scales, byte selectors, and protected BF16 tensors.
The frozen checkpoint-only protocol selected the Week-1 block64 model at
0.929 versus this model's 0.918. We nevertheless chose this artifact after it
won the sealed 2026 holdout 0.603 versus 0.492, with 8/1 paired flips
(p=0.039062). This is documented as a post-holdout robustness decision, not
as the precommitted winner.
For a full rebuild, train the short source with
training/reproduce_source.sh, set EAIMATH_BLOCK64_SOURCE to it and
EAIMATH_BLOCK64_REPORT to the downloaded block_adaptive_report.json, then
run convert_from_hf_checkpoint.py. Exact results are in results/.