Add IVF index for vec0 virtual table

Add inverted file (IVF) index type: partitions vectors into clusters via
k-means, quantizes to int8, and scans only the nearest nprobe partitions at
query time. Includes shadow table management, insert/delete, KNN integration,
compile flag (SQLITE_VEC_ENABLE_IVF), fuzz targets, and tests. Removes
superseded ivf-benchmarks/ directory.
This commit is contained in:
Alex Garcia 2026-03-29 19:46:23 -07:00
parent 43982c144b
commit 3358e127f6
22 changed files with 5237 additions and 28 deletions

View file

@ -93,13 +93,40 @@ $(TARGET_DIR)/rescore_quantize_edge: rescore-quantize-edge.c $(FUZZ_SRCS) | $(TA
$(TARGET_DIR)/rescore_interleave: rescore-interleave.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) -DSQLITE_VEC_ENABLE_RESCORE $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_create: ivf-create.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_operations: ivf-operations.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_quantize: ivf-quantize.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_kmeans: ivf-kmeans.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_shadow_corrupt: ivf-shadow-corrupt.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_knn_deep: ivf-knn-deep.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_cell_overflow: ivf-cell-overflow.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
$(TARGET_DIR)/ivf_rescore: ivf-rescore.c $(FUZZ_SRCS) | $(TARGET_DIR)
$(FUZZ_CC) $(FUZZ_CFLAGS) $(FUZZ_SRCS) $< -o $@
FUZZ_TARGETS = vec0_create exec json numpy \
shadow_corrupt vec0_operations scalar_functions \
vec0_create_full metadata_columns vec_each vec_mismatch \
vec0_delete_completeness \
rescore_operations rescore_create rescore_quantize \
rescore_shadow_corrupt rescore_knn_deep \
rescore_quantize_edge rescore_interleave
rescore_quantize_edge rescore_interleave \
ivf_create ivf_operations \
ivf_quantize ivf_kmeans ivf_shadow_corrupt \
ivf_knn_deep ivf_cell_overflow ivf_rescore
all: $(addprefix $(TARGET_DIR)/,$(FUZZ_TARGETS))