David S. Batista

icon indicating a website link www.davidsbatista.net icon indicating an email [email protected]

icon company @deepset-ai icon location Berlin, Germany icon location Software Developer | NLP Engineer | Machine Learning @deepset-ai

Results 94 comments of David S. Batista

`collection.iterator` created with `return_properties` fails when using properties with `object[]` `dataType`

Hi @tsmith023! Did you have the chance to work on this issue?

fix: window_size set during run instead of construction

@alperkaya I've reverted some of the changes you did, and added a few more tests - thanks for your contribution.

fix: window_size set during run instead of construction

@silvanocerza applied your changes request

feat: adding an `HybridRetriever` as a `Supercomponent` having `OpenSearch` as the document store

For the tests to pass, we will have to wait for a new release of haystack with the `TextEmbedder` protocol. But @sjrl feel free to give a second review

Document Store - closing resources

This was already previously discussed for Weavite https://github.com/deepset-ai/haystack-core-integrations/issues/471

Document Store - closing resources

to release async connections/resources the equivalent is `__aexit__`

Client Reject Incompatible models

both `NvidiaDocumentEmbedder` and `NvidiaTextEmbedder` define the backend as: ```python self.backend: Optional[Any] = None ``` while the `NvidiaGenerator` defines it as ```python self._backend: Optional[NimBackend] = None ``` maybe we can normalise...

Client Reject Incompatible models

```python class NvidiaGenerator: """ Generates text using generative models hosted with [NVIDIA NIM](https://ai.nvidia.com) on on the [NVIDIA API Catalog](https://build.nvidia.com/explore/discover) ``` typo above, double "on"

Client Reject Incompatible models

what about renaming: `integrations/nvidia/src/haystack_integrations/utils/nvidia/statics.py` to `model.py` since all the code on that file is related with Model information?

Client Reject Incompatible models

both `NvidiaDocumentEmbedder` and `NvidiaTextEmbedder` have the backend as: ```python self.backend: Optional[Any] = None ``` while the `NvidiaGenerator`: as ```python self._backend: Optional[Any] = None ``` let's have them all either as...