3 people 8ec1a07bfb feat: TurboQuant, KV cache compressor, utils, distortion suite #6 #7 #8 #10 ...

TurboQuant (Algorithm 2): PolarQuant(b-1) + QJL(1) = b bits total.
19 tests, 93% coverage.

KVCacheCompressor: TurboQuant for K cache (IP), TurboQuantMSE for V cache (MSE).
5 tests, 100% coverage.

Utils: bit packing, memory footprint calculation.
9 tests, 98% coverage.

Distortion validation suite: MSE bounds, IP bounds, monotonicity.
20 tests verifying against paper Table 2.

Full suite: 127 tests, 98% total coverage.

Codex review on TurboQuant:
- Noted that double-side IP estimation is biased (documented, tested)
- True bit rate is b + 32/d (norm overhead documented)
- QJL correction factor is inside QJL.dequantize (verified)

Co-Authored-By: tturney@psyguard.ai
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-24 21:04:56 -05:00

..

__init__.py

infra: project scaffold — CI, pyproject, CLAUDE.md, PLAN.md #1

2026-03-24 20:43:16 -05:00

test_codebook.py

feat: codebook construction for PolarQuant (codebook.py) #3

2026-03-24 21:00:07 -05:00

test_distortion.py

feat: TurboQuant, KV cache compressor, utils, distortion suite #6 #7 #8 #10

2026-03-24 21:04:56 -05:00

test_kv_cache.py

feat: TurboQuant, KV cache compressor, utils, distortion suite #6 #7 #8 #10

2026-03-24 21:04:56 -05:00

test_polar_quant.py

feat: PolarQuant quantizer + QJL 1-bit quantizer #4 #5

2026-03-24 21:02:06 -05:00

test_qjl.py

feat: PolarQuant quantizer + QJL 1-bit quantizer #4 #5

2026-03-24 21:02:06 -05:00

test_rotation.py

feat: random rotation matrix generation (rotation.py) #2

2026-03-24 20:53:52 -05:00

test_turboquant.py

feat: TurboQuant, KV cache compressor, utils, distortion suite #6 #7 #8 #10

2026-03-24 21:04:56 -05:00

test_utils.py

feat: TurboQuant, KV cache compressor, utils, distortion suite #6 #7 #8 #10

2026-03-24 21:04:56 -05:00