added encoder to whisper function in LLMWhisperClient

[ShoaibMajidDar]

When processing documents in Arabic, the expected Arabic text was not returned correctly.
I added an encoder argument to the whisper function in LLMWhisperClient. The issue was resolved by encoding the response in UTF-8, which correctly handled the Arabic text. The encoder is set to default to ISO-8859-1, but can now be adjusted as needed.

[hari-kuriakose]

@ShoaibMajidDar Thanks for the contribution!
This would really help. Just couple of minor suggestions and rest looks good.

[jaseemjaskp]

LGTM

[chandrasekharan-zipstack]

Ideally we need to forward the encoding in the request headers itself, so that it is understood by LLMWhisperer itself and is handled subsequently by the requests library. The current approach would help handle UTF-8 correctly which should cover most of the usecases and any requirement to support different encoding schemes will be properly tackled in the client and server in the future

[hari-kuriakose]

@chandrasekharan-zipstack Agree, let's take up the improvements as required later.