-
I have tried #8223 on a ~3.4TB gzipped Parquet dataset.
I tried four runs so far, with two different behaviours
- First I tried the whole dataset. I got to the last step (`to_parquet`), but then r…
-
**Describe the bug**
If the training data does not live on NFS but on node-specific storage, the current logic in https://github.com/NVIDIA/Megatron-LM/blob/0bc3547702464501feefeb5523b7a17e591b21fa/m…
-
### What happened?
``` Bash
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
Cell In[4], line…
-
### Preflight Checklist
- [X] I agree to follow the [Code of Conduct](https://github.com/dexidp/dex/blob/master/.github/CODE_OF_CONDUCT.md) that this project adheres to.
- [X] I have searched the [is…
-
**Is your feature request related to a problem? Please describe.**
If we could let Milvus support to save data in a distributed file system, that's will be convenient to use/save huge data.
**Desc…
-
An exception is thrown when loading parquet files into a ddf from Azure storage account. This only happens when using a distributed client, and it doesn't happen when using a local filesystem. The sta…
-
**What would you like to be added**:
1、Support `immediate` volumebindingmode storageclass
2、Support shared pvc used by multiple pods
**Why is this needed**:
1、Some commonly used distributed…
-
Hi,
I have a single GPU on my system and I am using CodeLlama-7b to test my environment.
I am running into the following error when I run the sample.
```
$ torchrun --nproc_per_node 1 example_co…
-
**Is your feature request related to a problem? Please describe.**
The problem is that the management component lacks high availability (HA) support. Currently, the management component is central to…
JaSei updated
3 weeks ago
-
### Checklist
- [X] 1. I have searched related issues but cannot get the expected help.
- [X] 2. The bug has not been fixed in the latest version.
### Describe the bug
使用最新版的lmdeploy 0.5.1在多卡V100或者…