deepgram-python-sdk
deepgram-python-sdk copied to clipboard
deepgram
Official Python SDK for Deepgram's automated speech recognition APIs.
## What is the current behavior? The UtteranceEnd event does not come 1000ms after the last spoken word. ## Steps to reproduce ``` options: LiveOptions = LiveOptions( model="nova-2", language="en", #...
## Description I'm using v2.11.0 of your SDK and I often run into this error in live transcriptions. ``` future: Traceback (most recent call last): File "/usr/local/lib/python3.11/dist-packages/deepgram/transcription.py", line 222, in...
## Proposed changes ## Context ## Possible Implementation ## Other information https://developers.deepgram.com/reference/create-key
## Proposed changes We should just require users to directly instantiate the classes and no longer have the Dot Notation mechanism. This is more or less a Javascript construct that...
## Proposed changes Currently, all WebSockets and HTTP(s) options are combined into a single ClientOptions class. You run into situations when options like `KeepAlive`, which are used for WebSockets, are...
## What is the current behavior? When installing on a system using Python version 3.9 or lower, the install mechanism should error out. There was an attempt to try an...
## Issue * Performing Latency calculations are a challenge. * Websocket reconnect so users don't need to redefine their events handlers in their code * Websocket native keep alive when...
## Issue Performing retries on failed Rest requests could be easier. ## Acceptance Criteria * For Rest Requests make it easy to implement retries.
## Issue * Specifically related to TTS implementing Chunking is a bit challenging for users. ## Acceptance Criteria * Chunking by having a native interface for different LLM responses *...
## Proposed changes Add Support for `no_delay` ## Context NA ## Possible Implementation NA ## Other information NA