Skip to content
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

update index.md #2965

Merged
merged 1 commit into from
Jan 6, 2025
Merged

update index.md #2965

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
13 changes: 7 additions & 6 deletions site/en/reference/index.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -409,15 +409,16 @@ Combined with SQ, HNSW_SQ offers a controllable trade-off between index size and
| `M` | M defines tha maximum number of outgoing connections in the graph. Higher M leads to higher accuracy/run_time at fixed ef/efConstruction. | [2, 2048] | None |
| `efConstruction` | ef_construction controls index search speed/build speed tradeoff. Increasing the efConstruction parameter may enhance index quality, but it also tends to lengthen the indexing time. | [1, int_max] | None |
| `sq_type` | Scalar quantizer type. | `SQ6`,`SQ8`, `BF16`, `FP16` | `SQ8` |
| `refine` | Whether refined data is reserved during index building. | `true`, `false` | `false` |
| `refine_type` | The data type of the refine index. | `SQ6`, `SQ8`, `BF16`, `FP16`, `FP32` | None |

- Search parameters

| Parameter | Description | Range | Default Value |
|---------------|------------------------------------------------------------------------------------------------------------|--------------------------------------|---------------|
| `ef` | Parameter controlling query time/accuracy trade-off. Higher `ef` leads to more accurate but slower search. | [`top_k`, int_max] | None |
| `refine` | Whether the refine is used during the train. | `true`, `false` | `false` |
| `refine_k` | The magnification factor of refine compared to *k*. | [1, *float_max*) | `1` |
| `refine_type` | The data type of the refine index. | `SQ6`, `SQ8`, `BF16`, `FP16`, `FP32` | None |


### HNSW_PQ

Expand All @@ -433,15 +434,15 @@ Combined with PQ, HNSW_PQ offers a controllable tradeoff between index size and
| `efConstruction` | ef_construction controls index search speed/build speed tradeoff. Increasing the efConstruction parameter may enhance index quality, but it also tends to lengthen the indexing time. | [1, int_max] | None |
| `m` | The number of sub-vector groups to split the vector into. | [1, 65536] | 32 |
| `nbits` | The number of bits into which each group of sub-vectors is quantized. | [1, 24] | 8 |
| `refine` | Whether refined data is reserved during index building. | `true`, `false` | `false` |
| `refine_type` | The data type of the refine index. | `SQ6`, `SQ8`, `BF16`, `FP16`, `FP32` | None |

- Search parameters

| Parameter | Description | Range | Default Value |
|---------------|------------------------------------------------------------------------------------------------------------|--------------------------------------|---------------|
| `ef` | Parameter controlling query time/accuracy trade-off. Higher `ef` leads to more accurate but slower search. | [`top_k`, int_max] | None |
| `refine` | Whether the refine is used during the train. | `true`, `false` | `false` |
| `refine_k` | The magnification factor of refine compared to *k*. | [1, *float_max*) | `1` |
| `refine_type` | The data type of the refine index. | `SQ6`, `SQ8`, `BF16`, `FP16`, `FP32` | None |

### HNSW_PRQ

Expand All @@ -458,15 +459,15 @@ Combined with a Product Residual Quantizer (PRQ), HNSW_PRQ offers an even higher
| `m` | The number of sub-vector groups to split the vector into. | [1, 65536] | 32 |
| `nbits` | The number of bits into which each group of sub-vectors is quantized. | [1, 24] | 8 |
| `nrq` | The number of residual subquantizers. | [1, 16] | 2 |
| `refine` | Whether refined data is reserved during index building. | `true`, `false` | `false` |
| `refine_type` | The data type of the refine index. | `SQ6`, `SQ8`, `BF16`, `FP16`, `FP32` | None |

- Search parameters

| Parameter | Description | Range | Default Value |
|---------------|------------------------------------------------------------------------------------------------------------|--------------------------------------|---------------|
| `ef` | Parameter controlling query time/accuracy trade-off. Higher `ef` leads to more accurate but slower search. | [`top_k`, int_max] | None |
| `refine` | Whether the refine is used during the train. | `true`, `false` | `false` |
| `refine_k` | The magnification factor of refine compared to *k*. | [1, *float_max*) | `1` |
| `refine_type` | The data type of the refine index. | `SQ6`, `SQ8`, `BF16`, `FP16`, `FP32` | None |

</div>

Expand Down
Loading