haystack
haystack copied to clipboard
deepset-ai
:mag: AI orchestration framework to build customizable, production-ready LLM applications. Connect components (models, vector DBs, file converters) to pipelines or agents that can interact with your d...
Once the simplification of `haystack/modeling/model/language_modeling.py` are merged, we can proceed attempting to load a `Data2VecVision` into a Retriever, be it `EmbeddingRetriever` or a custom retriever class. This issue implies also...
**Describe the bug** In the past we were running benchmarks quite regularly for haystack and published results [here](https://haystack.deepset.ai/benchmarks/latest). As the benchmarking script was not working anymore at some point, we...
**Is your feature request related to a problem? Please describe.** [This line](https://github.com/deepset-ai/haystack/blob/632cd1c141a8b485c6ef8695685d2d8eef3ca50f/haystack/nodes/reader/farm.py#L209) sets the parameter `save_dir` to a directory two layers above the current directory. This can be confusing and...
Currently, Haystack supports storing data into ElasticSearch, InMemory, and RDBMS. It would be nice to add support of Object storage like S3, which is very cheap and have less hassle...
### Discussed in https://github.com/deepset-ai/haystack/discussions/2000 Originally posted by **jacoby149** January 12, 2022 Hi, I would like to do population based training to train my farm reader. I know haystack uses ray...
**Problem:** I want to retrieve all relevant (similar) documents from the `ElasticsearchDocumentStore` based on the `_score` using the `EmbeddingRetriever` (I am not using the Reader). Prior to the search, I...
Seems like Milvus will introduce an Embedded version: - https://github.com/deepset-ai/haystack/issues/2081#issuecomment-1066044687 - https://wiki.lfaidata.foundation/display/MIL/MEP+26+--+Embedded+Milvus - https://github.com/milvus-io/milvus/issues/15711 It's probably worth keeping an eye on it to see if it can simplify drastically our...
As suggested by @vblagoje on the Haystack slack, it would be nice to have a helper function, similar to `open_search_index_to_documentstore` or `convert_files_to_docs` that would allow users to provide a HF...
**Problem** Currently the saved node output from `pipe.eval()` is in csv. This makes it quite annoying, as it is never configured right to have a look into it right away....
Currently, the no_answer scores use an `expit(score/8)` function to be scaled to the interval 0 to 1. However, the score should be between 0 and 1 even before that because...
Metadata
Owner
Metadata
:mag: AI orchestration framework to build customizable, production-ready LLM applications. Connect components (models, vector DBs, file converters) to pipelines or agents that can interact with your d...