-
nProbe cento v.1.5.180316
This is my cento command:
```
cento -i ens2f1 --zmq tcp://172.XX.XX.XX:5556
cento -i ens2f0 --zmq tcp://172.XX.XX.XX:5556
```
This is my ntopng command:
```
ntopng …
-
I use IndexHNSWFlat make Index.
```
hnsw_index = faiss.IndexHNSWFlat(128, 32, faiss.METRIC_INNER_PRODUCT)
hnsw_index.hnsw.efConstruction = 64
```
```
hnsw_index = faiss.read_index('faiss.index'…
-
I'm trying to adjust the learning rates for training with 2 GPUs by first trying the same learning rates used for 8 GPUs but also dividing all learning rates by 4 in proportion to only using 2 GPUs. A…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Describe the bug
I am unable to connect.
`connections.connect(host="localhost", port="19530")`
``…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Environment
```markdown
- Milvus version: 2.4-20240412-9613d368-amd64
- Deployment mode(standalone or cluster…
-
Please check if all columns can be sorted
-
Hi,用了如下的建表语句,数据2亿,召回1000个。结果只有80%与暴力检索的结果一致,请问为什么这么低呢?
{
"name": "piv_vec",
"partition_num": 1,
"replica_num": 1,
"engine": {
"name": "gamma",
"index_size": 3276800,
"ret…
-
**Describe the bug**
I am running [step 3](https://github.com/NVIDIA/Megatron-LM/blob/InstructRetro/tools/retro/build_db.md#step-3-build-index-for-similarity-search) on one 80G A100 GPU to "Build ind…
-
# Summary
If `faiss` is `import`ed before `torch` - the returned KNNs from a GPU index are very bad. If we switch the order of imports, the returned KNNs are **much** more accurate.
# Platform
…
-
Vearch 3.2.5版本
IVFPQ索引使用SIFT1M数据集自己进行的召回率测试仅有57,和官网召回率测试结果有较大出入,不知道问题出在哪里
下面是我的测试结果
ncen | nsub | nprobe | recall@100
-- | -- | -- | --
256 | 32 | 1 | 33.3205
256 | 32 | 5 | 52.5332
256 | 32 | …