Mikko Lehtimäki
Mikko Lehtimäki
If I query against GPTSimpleVectorIndex, the response has source nodes that can be looked at to determine the original document id. Very useful behavior. If I query against GPTQdrantIndex, source...
Small fix for getting data from Qdrant. Now `doc_id` is correctly found from the Qdrant payload. Discussed at https://github.com/jerryjliu/gpt_index/issues/449
When creating a ComposableGraph over a composite index of Qdrant indices, queries error in File ...gpt_index/gpt_index/indices/query/vector_store/qdrant.py:61, in GPTQdrantIndexQuery.__init__(self, index_struct, embed_model, similarity_top_k, client, * *kwargs) 58 raise ValueError(import_err_msg) 60 if client...
A request for adding basic vector arithmetic on the Qdrant cluster. These would save users from loading hundreds of vectors and implementing computation on client side. API could be like...
Trying to save optimization results using the `save()` method will crash on `TypeError: is not JSON serializable`.
Cannot seem to find a way to do this. The `init_df` accepts a logging level argument, but it doesn't seem to prevent logging if I set my loggers to a...
Seems that the default value is `-15`. When trying to set a negative value, the following error is shown: ``` deep-filter --output-dir out-deepfilter -a 10 --min-db-thresh '-14' file.wav error: unexpected...
So calling `init_df` gives me the model and a `state`. If I want to make multiple calls to `enhance`, can I keep using the same `state`? Calling `state.reset()` does not...
Typically it's possible to load torch models to cpu / gpu by using the `map_location` argument. This doesn't work for the acoustic model: `TypeError: hubert_soft() got an unexpected keyword argument...