SambaNova STT

The SambaNovaSTTService provides high-efficiency inference for real-time speech transcription using SambaNova's optimized Whisper API.

Installation

To use SambaNova, install the required dependencies:

pip install "piopiy-ai[sambanova]"

Prerequisites

A SambaNova account and API key (Get yours here).

Set your API key in your environment:

export SAMBANOVA_API_KEY="your_api_key_here"

Configuration

`SambaNovaSTTService` Parameters

Parameter	Type	Default	Description
`api_key`	`str`	Required	Your SambaNova API key.
`model`	`str`	`"Whisper-Large-v3"`	The Whisper model to use.
`base_url`	`str`	`"https://api.sambanova.ai/v1"`	SambaNova API base URL.
`language`	`Language`	`Language.EN`	Language of the audio input.
`prompt`	`str`	`None`	Optional text to guide the model's style.
`temperature`	`float`	`0.0`	Sampling temperature between 0 and 1.

Usage

Basic Setup

import os
from piopiy.services.sambanova.stt import SambaNovaSTTService

stt = SambaNovaSTTService(
    api_key=os.getenv("SAMBANOVA_API_KEY")
)

With Custom Model and Language

import os
from piopiy.services.sambanova.stt import SambaNovaSTTService
from piopiy.transcriptions.language import Language

stt = SambaNovaSTTService(
    api_key=os.getenv("SAMBANOVA_API_KEY"),
    model="Whisper-Large-v3",
    language=Language.EN_US
)

Notes

Optimized Inference: SambaNova's platform is engineered for extremely fast execution of foundation models like Whisper.
Probability Metrics: Note that SambaNova STT does not currently support probability metrics.

Installation​

Prerequisites​

Configuration​

SambaNovaSTTService Parameters​

Usage​

Basic Setup​

With Custom Model and Language​

Notes​