tony comments

Results 26 comments of


                                            tony

Sqlalchemy 2 cannot be used in projects using langchain.

+1 this. It would be great if sqlalchemy was an optional dependency since (1) we dont use it's features in langchain and (2) it conflicts with our existing codebase which...

PBR Materials Support

Hm, with separate shaders as currently implemented, I wonder if there can be an easy way to load meshes with predefined material data (i.e. like in a glb). Could be...

Add example to consume streams

Any updates on this? Is there a way to get the raw stream from the MediaStream currently?

feat(python): support merge_columns via Ray integration

Any updates on this?

Out of Memory when using Streaming Dataloader

To echo @VikaasVarma's point here-- the copy on read issues with the torch dataloader comes back to the dataset object storing a large number of naive python objects that can't...

[add_columns] implements add columns task through ray

Also interested in this. I believe this would just require exposing `add_columns_to_fragments` in python to parallelize the fragment updates and updating the `LanceCommitter` to handle the transaction type

feat: add distribute add columns by ray

Is there a reason why this can't be as simple as https://github.com/dream3d-ai/lance/pull/1/files

feat: add distribute add columns by ray

> > Is there a reason why this can't be as simple as https://github.com/dream3d-ai/lance/pull/1/files > > @tonyf If there is only one add_column, your implementation seems to be fine. However,...

lance.torch.data.LanceDataset + torch.utils.data.DataLoader is slower than to_batches + frombuffer

This difference here is likely due to the fact your test is reading the dataset in order without any sharding. This means you're getting the added benefit of batch _and_...