Skip to content

[Feature] Add min_nprobe to get_optimal_hyperparameters function #174

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
victor-paltz wants to merge 1 commit into
base: master
Choose a base branch
from
+13 −5
Open

[Feature] Add min_nprobe to get_optimal_hyperparameters function #174

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 13 additions & 5 deletions autofaiss/external/optimize.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -140,8 +140,8 @@ def get_optimal_index_keys_v2(
nb_vectors: int,
dim_vector: int,
max_index_memory_usage: str,
flat_threshold: int = 1000,
quantization_threshold: int = 10000,
flat_threshold: int = 1_000,
quantization_threshold: int = 10_000,
force_pq: Optional[int] = None,
make_direct_map: bool = False,
should_be_memory_mappable: bool = False,
Expand Down Expand Up @@ -477,6 +477,8 @@ def get_optimal_hyperparameters(
max_timeout_per_iteration_s: float = 1.0, # seconds
min_ef_search: int = 32,
min_nearest_neighbors_to_retrieve: int = 20,
*,
min_nprobe: int = 1,
) -> str:
"""Find the optimal hyperparameters to maximize the recall given a query speed in milliseconds/query"""

Expand All @@ -487,15 +489,19 @@ def get_optimal_hyperparameters(
ht = 2048
nb_clusters = int(params[2])
hyperparameter_str_from_param = lambda nprobe: f"nprobe={nprobe},ht={ht}"
parameter_range = list(range(1, min(6144, nb_clusters) + 1))
nprobe_lower_bound = max(1, min(min_nprobe, nb_clusters))
nprobe_upper_bound = max(nprobe_lower_bound, min(6144, nb_clusters))
parameter_range = list(range(nprobe_lower_bound, nprobe_upper_bound + 1))
timeout_boost_for_precision_search = 6.0

elif any(re.findall(r"OPQ\d+_\d+,IVF\d+_HNSW\d+,PQ\d+", index_key)):

ht = 2048
nb_clusters = int(params[2])
hyperparameter_str_from_param = lambda nprobe: f"nprobe={nprobe},efSearch={2*nprobe},ht={ht}"
parameter_range = list(range(max(1, min_ef_search // 2), min(6144, nb_clusters) + 1))
nprobe_lower_bound = min(nb_clusters, max(min_nprobe, min_ef_search // 2, 1))
nprobe_upper_bound = max(nprobe_lower_bound, min(6144, nb_clusters))
parameter_range = list(range(nprobe_lower_bound, nprobe_upper_bound + 1))
timeout_boost_for_precision_search = 12.0

elif any(re.findall(r"HNSW\d+", index_key)):
Expand All @@ -508,7 +514,9 @@ def get_optimal_hyperparameters(

nb_clusters = int(params[0])
hyperparameter_str_from_param = lambda nprobe: f"nprobe={nprobe}"
parameter_range = list(range(1, nb_clusters + 1))
nprobe_lower_bound = min(min_nprobe, nb_clusters)
nprobe_upper_bound = nb_clusters
parameter_range = list(range(nprobe_lower_bound, nprobe_upper_bound + 1))
timeout_boost_for_precision_search = 6.0

elif index_key == "Flat":
Expand Down