c111b3a9b0
Phase 01 (LLM Tuning): - Gemma4 26B: 74.65 t/s (fast) - Qwen 35B: 61.62 t/s (balanced) - Gemma4 31B: 16.0 t/s (deep-coder) - Qwen 27B: 16.7 t/s (deep-logic) - Qwen 122B: 8.95 t/s (ultra, GPU 1 only) Phase 02 (API Engine): - FastAPI reverse proxy on port 8000 - /engine/switch hot-swap with 503 protection - config/engine_models.json as single source of truth - Replaced 4 individual .bat files with unified engine File cleanup: - scripts/ 85 files -> 9 + _archive/ - Root .bat files -> _archive/
46 lines
1.3 KiB
Python
46 lines
1.3 KiB
Python
"""Quick benchmark for running llama-server instance"""
|
|
import urllib.request, json, time, sys
|
|
|
|
BASE = "http://127.0.0.1:8000"
|
|
RUNS = 5
|
|
TOKENS = 200
|
|
|
|
def bench(max_tokens=TOKENS):
|
|
payload = json.dumps({
|
|
"model": "m",
|
|
"messages": [{"role": "user", "content": "Count from 1 to 100, each number on a new line."}],
|
|
"max_tokens": max_tokens,
|
|
"temperature": 0
|
|
}).encode()
|
|
req = urllib.request.Request(
|
|
f"{BASE}/v1/chat/completions",
|
|
data=payload,
|
|
headers={"Content-Type": "application/json"}
|
|
)
|
|
t0 = time.time()
|
|
resp = json.loads(urllib.request.urlopen(req, timeout=300).read())
|
|
dt = time.time() - t0
|
|
ct = resp.get("usage", {}).get("completion_tokens", 0)
|
|
return ct / dt if dt > 0 else 0, ct, dt
|
|
|
|
print("Warmup...", flush=True)
|
|
try:
|
|
bench(20)
|
|
except Exception as e:
|
|
print(f"Warmup failed: {e}")
|
|
sys.exit(1)
|
|
print("Warmup done\n", flush=True)
|
|
|
|
speeds = []
|
|
for i in range(RUNS):
|
|
tps, ct, dt = bench()
|
|
speeds.append(tps)
|
|
print(f" Run {i+1}: {tps:.2f} t/s (tokens={ct}, time={dt:.2f}s)", flush=True)
|
|
|
|
avg = sum(speeds) / len(speeds)
|
|
best = max(speeds)
|
|
mn = min(speeds)
|
|
print(f"\n{'='*50}")
|
|
print(f" RESULT: AVG {avg:.2f} / BEST {best:.2f} / MIN {mn:.2f} t/s")
|
|
print(f"{'='*50}")
|