-
| | db | table | e.g. |
|---|----|-------|------------------------------------------------------------------|
| | 1 | 1 | tb_user…
-
1. The API for shardingDefinition and shardingLifeCycleAction was introduced in 1.0, however shardingLifeCycleAction has not been implemented yet, including sharding preterminate, postprovision, shard…
-
Hi Torchrec Team,
I'm following the protocol in https://github.com/pytorch/torchrec/blob/main/TorchRec_Interactive_Tutorial_Notebook_OSS_version.ipynb to setup sharded quantization tables, but have…
-
### 🐛 Describe the bug
I have noticed `test_linear_row_wise_parallel` fails when run with 6 GPUs
https://github.com/pytorch/pytorch/blob/f2f7ef9d5908f53a5fd7991ed0a1ef99069813a1/test/distribute…
-
sharding version: 5.1.1
We found some problems here. We use gorm to link sharding and sharding to access mysql. Currently, we find that sharding has been growing CLOSE_WAIT links. At the same time, s…
-
List post #6286 that aren't breaking but would be nice to have
```[tasklist]
### Tasks
- [ ] Functioning Test Report
- [ ] Smart number of shards
- [ ] Avoid failing conda tests
- [ ] Add branch chec…
-
-
Hi !
I was playing with JAX on Neuron recently and came across a bug that is quite annoying.
When trying to shard a very simple MLP layer, depending on the axis you choose, the compilation fails…
-
### Is there an existing issue for the same bug?
- [X] I have checked the existing issues.
### Branch Name
main
### Commit ID
b0855d26f
### Other Environment Information
```Markdown
- Hardware …
-
**Description:**
Hi ,torchrec team,
I'm using EmbeddingCollection and constrain the sharding type as DATA_PARALLEL. Subsequently, I should be able to get the parameters and pass it to my optimizers…