Turn Off return of the verbose unstructured text from --> result = RAG.search(query=query,k=4)

How to turn off return of -- the verbose unstructured text -->
using --> result = RAG.search(query=query,k=4)
and just Return the dict -> [{"content": ...

[Aug 18, 16:57:05] Loading segmented_maxsim_cpp extension (set COLBERT_LOAD_TORCH_EXTENSION_VERBOSE=True for more info)...
Loading searcher for index ragcorpus for the first time... This may take a few seconds
[Aug 18, 16:57:06] #> Loading codec...
[Aug 18, 16:57:06] #> Loading IVF...
[Aug 18, 16:57:06] Loading segmented_lookup_cpp extension (set COLBERT_LOAD_TORCH_EXTENSION_VERBOSE=True for more info)...
[Aug 18, 16:57:06] #> Loading doclens...
[Aug 18, 16:57:06] #> Loading codes and residuals...
[Aug 18, 16:57:06] Loading filter_pids_cpp extension (set COLBERT_LOAD_TORCH_EXTENSION_VERBOSE=True for more info)...
[Aug 18, 16:57:06] Loading decompress_residuals_cpp extension (set COLBERT_LOAD_TORCH_EXTENSION_VERBOSE=True for more info)...
Searcher loaded!

#> QueryTokenizer.tensorize(batch_text[0], batch_background[0], bsize) ==
#> Input: . who is Forneus?, True, None
#> Output IDs: torch.Size([32]), tensor([ 101, 1, 2040, 2003, 2005, 2638, 2271, 1029, 102, 103, 103, 103,
103, 103, 103, 103, 103, 103, 103, 103, 103, 103, 103, 103,
103, 103, 103, 103, 103, 103, 103, 103])
#> Output Mask: torch.Size([32]), tensor([1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0])

***
I just want this -->
[{"content": ...

Start at the RAG.search(query=query, k=4) entry point and reproduce the verbose loading and tokenizer output shown in the issue. Determine how output is produced and whether it can be disabled while preserving the requested [{"content": ...}] result; done means the search returns that result without the unwanted verbose text.

Written by the indexing model from the issue text.