6cfcbc6bc0
vllm/gemma-31b-dual now serves cyankiwi QAT-AWQ-int4 + bf16 KV @224K on stock vLLM v0.24.0 (base.yml), OVERLAY-FREE — folds onto vllm-stable, retiring the 31b's vllm-gemma-stable dependence. Promoted to ⚠️ Production w/ caveats (validated 2026-07-02: verify-full 9/9, verify-stress→210K @ 1162MB VRAM margin, bench ~59 TPS, soak PASS). bf16 not int8-PTH: on v0.24.0 int8-PTH allocates 262K but silently craters recall past ~32K (needs PR #40391, open/unmerged upstream — the same cyankiwi weights recall clean to 112K+ on v0.22.0+#40391). int8-PTH 262K returns free when #40391 merges. MTP disabled (Gemma-4 MTP x tools broken on v0.24.0, vLLM #39043 / #42006 closed-unmerged). - remove the broken #537 int8.yml (recall cliff); base.yml replaces it - DEFAULTS (gemma-4-31b,vllm,dual) -> vllm/gemma-31b-dual - deprecate gemma-int8-mtp / gemma-bf16-mtp / qat-w4a16 (registry + header) - fix 2 false "gemma int8-PTH native @262K" claims in vllm-stable.yml - repoint launch/preflight/setup/switch hints off the deprecated slugs - update resolver + setup-picker fixtures; add BENCHMARKS row Suite: 59 pass / 1 pre-existing (test-submit-bench worktree-fixture-absent). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EfF565T9eSLaqGzidyJ1Pm
65 lines
3.3 KiB
Bash
Executable File
65 lines
3.3 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
# PR-B — <engine>/default resolver uses DEFAULTS + detected topology.
|
|
set -euo pipefail
|
|
|
|
ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
|
|
cd "$ROOT_DIR"
|
|
|
|
assert_contains() {
|
|
local haystack="$1" needle="$2"
|
|
if [[ "$haystack" != *"$needle"* ]]; then
|
|
echo "ASSERTION FAILED: expected output to contain: $needle" >&2
|
|
echo "--- output ---" >&2
|
|
echo "$haystack" >&2
|
|
exit 1
|
|
fi
|
|
}
|
|
|
|
fake_one='0:RTX_3090:24576:8.6'
|
|
fake_two='0:RTX_3090:24576:8.6,1:RTX_3090:24576:8.6'
|
|
|
|
out="$(CLUB3090_FAKE_GPUS="$fake_one" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection vllm/default 2>&1)"
|
|
assert_contains "$out" "selected variant: vllm/minimal"
|
|
assert_contains "$out" "vllm/minimal"
|
|
|
|
out="$(CLUB3090_FAKE_GPUS="$fake_two" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection vllm/default 2>&1)"
|
|
assert_contains "$out" "selected variant: vllm/dual"
|
|
assert_contains "$out" "vllm/dual"
|
|
|
|
out="$(CLUB3090_FAKE_GPUS="$fake_one" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection vllm/dual/default 2>&1)"
|
|
assert_contains "$out" "selected variant: vllm/dual"
|
|
|
|
if out="$(CLUB3090_FAKE_GPUS="$fake_one" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection llamacpp/dual/default 2>&1)"; then
|
|
echo "ASSERTION FAILED: bad topology default unexpectedly resolved" >&2
|
|
echo "$out" >&2
|
|
exit 1
|
|
fi
|
|
assert_contains "$out" "cannot resolve default variant 'llamacpp/dual/default'"
|
|
assert_contains "$out" "Available defaults"
|
|
|
|
out="$(NVIDIA_VISIBLE_DEVICES=0,1 FORCE=1 PREFLIGHT_NO_COMPOSE_DEPS=1 COMPOSE_BIN=: READY_TIMEOUT=1 bash scripts/switch.sh --no-wait vllm/default 2>&1 || true)"
|
|
assert_contains "$out" "bringing up: vllm/dual"
|
|
|
|
out="$(NVIDIA_VISIBLE_DEVICES=0 FORCE=1 PREFLIGHT_NO_COMPOSE_DEPS=1 COMPOSE_BIN=: READY_TIMEOUT=1 bash scripts/switch.sh --no-wait vllm/dual/default 2>&1 || true)"
|
|
assert_contains "$out" "bringing up: vllm/dual"
|
|
|
|
# PR-B: `<model>/default` token dispatch through launch.sh (engine-vs-model).
|
|
# Single rig: qwen3.6-27b/default → curated beellama (ENGINE_PREFERENCE #1, caveats);
|
|
# dual → vllm/dual.
|
|
out="$(CLUB3090_FAKE_GPUS="$fake_one" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection --variant qwen3.6-27b/default 2>&1)"
|
|
assert_contains "$out" "selected variant: beellama/dflash"
|
|
out="$(CLUB3090_FAKE_GPUS="$fake_two" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection --variant qwen3.6-27b/default 2>&1)"
|
|
assert_contains "$out" "selected variant: vllm/dual"
|
|
# gemma-4-31b/default dual → vllm/gemma-31b-dual (model token overrides PRIMARY_MODEL).
|
|
out="$(CLUB3090_FAKE_GPUS="$fake_two" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection --variant gemma-4-31b/default 2>&1)"
|
|
assert_contains "$out" "selected variant: vllm/gemma-31b-dual"
|
|
# Unknown X/default → clear error (neither engine nor model).
|
|
if out="$(CLUB3090_FAKE_GPUS="$fake_one" SWITCH=/bin/echo bash scripts/launch.sh --no-preflight --no-verify --no-projection --variant bogus/default 2>&1)"; then
|
|
echo "ASSERTION FAILED: bogus/default unexpectedly resolved" >&2
|
|
echo "$out" >&2
|
|
exit 1
|
|
fi
|
|
assert_contains "$out" "neither a known engine nor a known model"
|
|
|
|
echo "test-default-resolver: ok"
|