Tool to provide voice input support for Ollama models.

Description: Added voice input tool for Ollama model that holds two classes that extend BaseBlobParser & BaseTool respectively, to allow for voice input (pre-saved or on the spot) on various Ollama model chains, as the user requires. Details on how to utilize the tool (e.g. use cases) can be found in the docs/integrations/tools folder in 'ollama_voice_input.ipynb'.

Issue: #27717

Dependencies: 'sounddevice', 'scipy' for voice recording (not anymore as per my knowledge, except for note below)

Twitter Handle: @faizannaseerr

Lint & Test: I ran lint & tests locally, and everything seemed fine as per my knowledge.

Added Test & Docs: Unit tests for modules and docs on how to use the tool have been added.

Additional Notes:

• There is a unit test for long audio input that should use an audio file named 'long_audio.wav' in the langchain_community/tests/examples file. I have not added this file yet as I could not find any long, free-to-use, audio input file online and therefore can leave this unit test out if required, or can use it if a long audio input file is found (that leads to several Document objects after the parse)
• Will fix imports shortly if required (still some being re-imported from LangChain, not sure if they need to be altered or not)