Phase 2.6.1: co_occurrence_edges migration - build_co_occurrence_edges_from_qdrant() - Qdrant embeddings → frame grouping → YOLO objects - Result: 6679 edges (vs 6701 PostgreSQL) Phase 2.6.2: face_face_edges migration - build_face_face_edges_from_qdrant() - Qdrant embeddings → frame grouping → face pairs - mutual_gaze detection preserved - Result: 6 edges (exact match) Phase 2.6.3: speaker_face_edges migration - build_speaker_face_edges_from_qdrant() - Qdrant embeddings → trace_id frame ranges - SPEAKS_AS edge creation Architecture: - All edges use Qdrant payload (no face_detections queries) - PostgreSQL fallback for empty Qdrant - Estimated 3.6x performance improvement Testing: - Playground (3003): ✓ All Phase 2.6 logs verified - Edge counts: ✓ Close match with PostgreSQL - Fallback: ✓ Working Docs: - docs_v1.0/DESIGN/TKG_PHASE2_6_EDGES_MIGRATION.md - docs_v1.0/M4_workspace/2026-06-21_phase2_6_test.md
43 lines
1.1 KiB
Bash
Executable File
43 lines
1.1 KiB
Bash
Executable File
#!/bin/bash
|
|
# Start Gemma 4 E4B with vMLX (Vision + Audio, 8-bit)
|
|
# Per: GEMMA4_E4B_4BIT_SETUP.md
|
|
|
|
VMLX="/Users/accusys/vmlx/.venv/bin/vmlx"
|
|
MODEL="/Users/accusys/models/mlx-gemma4-e4b-it-8bit"
|
|
PORT=8000
|
|
LOG_FILE="/Users/accusys/momentry_core/logs/vmlx_8000.log"
|
|
|
|
# Kill existing processes on port
|
|
lsof -i :$PORT | awk 'NR>1 {print $2}' | while read pid; do kill $pid; done
|
|
sleep 2
|
|
|
|
# Start vMLX server (editable install v1.5.59)
|
|
$VMLX serve $MODEL \
|
|
--host 0.0.0.0 --port $PORT \
|
|
--enable-prefix-cache \
|
|
--use-paged-cache \
|
|
--enable-disk-cache \
|
|
--kv-cache-quantization q8 \
|
|
--max-cache-blocks 2048 \
|
|
--timeout 1200 \
|
|
--log-level INFO \
|
|
--served-model-name gemma-4-E4B \
|
|
> $LOG_FILE 2>&1 &
|
|
|
|
echo "vMLX server starting on port $PORT"
|
|
echo "Model: Gemma 4 E4B 8bit (MLX) — supports Vision + Audio"
|
|
echo "Log: $LOG_FILE"
|
|
|
|
# Wait for ready
|
|
for i in $(seq 1 30); do
|
|
if curl -s -m 2 http://localhost:$PORT/health >/dev/null 2>&1; then
|
|
echo "✅ Ready ($i s)"
|
|
echo "API: http://localhost:$PORT/v1/chat/completions"
|
|
exit 0
|
|
fi
|
|
sleep 2
|
|
done
|
|
|
|
echo "❌ Not ready after 60s, check log"
|
|
exit 1
|