Qwen3-30B parity on M3U96a (bit-identical, 121.7 vs 121.5 tok/s), make_tables.py, cold-NVMe sweep script, LOG
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
23 changed files +2,727 −9
added
benchmarks/processed/gateC/sim_olmoe_v2_ssd043_m8.json
+2309 −0
Large diff (2,310 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/gateC/cold_nvme_laptop/sweep.jsonl
+3 −0
Large diff (3 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/gateC/qwen3_m3u96a/baseline.jsonl
+48 −0
Large diff (48 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/tests/m3u96a/greedy_qwen3_moe_cap32_20261101T014433.json
+14 −0
Large diff (15 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/tests/m3u96a/numerical_qwen3_moe_cap128_2026-11-01T014417.json
+50 −0
Large diff (51 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/tests/m3u96a/numerical_qwen3_moe_cap16_2026-11-01T014425.json
+50 −0
Large diff (51 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/tests/m3u96a/numerical_qwen3_moe_cap32_2026-11-01T014422.json
+50 −0
Large diff (51 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/tests/m3u96a/numerical_qwen3_moe_cap64_2026-11-01T014420.json
+50 −0
Large diff (51 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/raw/tests/m3u96a/pytest_qwen3_m3u96a_20261101T014415.txt
+2 −0
Large diff (2 lines) is not loaded by default to keep the page fast.Load diff
added
benchmarks/scripts/run_cold_nvme_laptop.sh
+17 −0
Large diff (17 lines) is not loaded by default to keep the page fast.Load diff
added
paper/main.pdf
+0 −0
Binary file not shown.
modified
paper/main.tex
+1 −1
Large diff (8 lines) is not loaded by default to keep the page fast.Load diff
modified
paper/sections/ablations.tex
+1 −1
Large diff (8 lines) is not loaded by default to keep the page fast.Load diff
modified
paper/sections/conclusion.tex
+1 −1
Large diff (8 lines) is not loaded by default to keep the page fast.Load diff
modified
paper/sections/formulation.tex
+1 −1
Large diff (8 lines) is not loaded by default to keep the page fast.Load diff
modified
paper/sections/implementation.tex
+2 −2
Large diff (10 lines) is not loaded by default to keep the page fast.Load diff
modified
paper/sections/results.tex
+2 −2
Large diff (16 lines) is not loaded by default to keep the page fast.Load diff
modified
paper/sections/setup.tex
+1 −1
Large diff (7 lines) is not loaded by default to keep the page fast.Load diff
added
paper/tables/cold.tex
+14 −0
Large diff (14 lines) is not loaded by default to keep the page fast.Load diff
added
paper/tables/qwen3.tex
+16 −0
Large diff (16 lines) is not loaded by default to keep the page fast.Load diff
added
paper/tables/sim.tex
+29 −0
Large diff (29 lines) is not loaded by default to keep the page fast.Load diff
modified
research/LOG.md
+9 −0
Large diff (12 lines) is not loaded by default to keep the page fast.Load diff
added
scripts/make_tables.py
+57 −0
Large diff (58 lines) is not loaded by default to keep the page fast.Load diff