Felipe Cardoso

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-04-15 13:55:38 +00:00

c847991740 docs: add agentic coding evaluation landscape research

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-04-12 23:12:35 +00:00

15bb6a8ed9 feat(serve): set APEX I-Compact as default, harden benchmark workflow

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-04-03 18:04:15 +00:00

474d94a07e chore: update model catalog with gemma 4, opus distill, and hw-bandwidth target

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-31 08:10:54 +00:00

6ab08537ca fix: address code review findings — batch args, venv path, serve flags

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-30 19:13:55 +00:00

dd403a907c feat(serve): add optimized llama-server launcher with n-gram speculation

ba24091791 feat(benchmark): add -b/--batch flag, test MoE batch size impact

ea70687cd2 docs: update optimization guide with measured hardware data

1549bc27c0 feat(optimize): add Phase 2 power profile and system tuning

f92b710492 fix(benchmark): parse llama-bench output with variable column count

Compare 6 commits »

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 21:48:33 +00:00

38daf953bf feat: add --pp and --tg flags for realistic benchmark workloads

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 21:46:16 +00:00

3686783f4d feat: add --context flag for configurable long-context benchmarks

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 21:38:49 +00:00

1b5b193e81 fix: suppress exit code 143 from metric logger cleanup

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 18:23:06 +00:00

fb1e57f1bf feat: make llama-rocm-7.2 a required toolbox in benchmark setup

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 18:17:16 +00:00

7c8be55bfe fix: resolve model paths for toolbox container access

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 18:14:31 +00:00

d22c062ca7 fix: model catalog shows download status, GPU detection in toolbox

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 18:11:00 +00:00

6f197a1455 fix: pass ARGS through in benchmark Makefile targets

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 18:07:27 +00:00

cb25fa3f6f feat: add benchmark filtering (--max-size, --category, --skip-longctx)

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-26 08:44:18 +00:00

eb52ea52ce fix: follow symlinks in model discovery, update model catalog

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-25 23:20:24 +00:00

58124cd657 feat: add Qwen3.5 model catalog and agentic evaluation framework

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-25 22:58:19 +00:00

71053997be chore: remove .idea from tracking, add to .gitignore

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-25 21:22:42 +00:00

e9cb5c491f fix+test: improve test suite, fix 2 bugs found by tests

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-25 21:15:36 +00:00

a403dd9ce0 test: add BATS test suite (79 tests)

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-25 20:44:16 +00:00

da2c4c6b8a fix(docs): address review findings — accuracy, consistency, completeness

cardosofelipe pushed to main at cardosofelipe/strix-halo-optimizations

2026-03-25 19:50:22 +00:00

5b81437637 docs: add README, CLAUDE.md, AGENTS.md, and full docs/ suite