• Stars
    star
    52,286
  • Rank 188 (Top 0.01 %)
  • Language
    Python
  • License
    MIT License
  • Created over 1 year ago
  • Updated 3 months ago

Reviews 5.0 (1)

4 months ago by ibrarmalik

Works like magic, and easy to set-up. I've used this to transcribe long podcasts and make them searchable.

Repository Details

Robust Speech Recognition via Large-Scale Weak Supervision

Whisper

[Blog] [Paper] [Model card] [Colab example]

Whisper is a general-purpose speech recognition model. It is trained on a large dataset of diverse audio and is also a multitasking model that can perform multilingual speech recognition, speech translation, and language identification.

Approach

Approach

A Transformer sequence-to-sequence model is trained on various speech processing tasks, including multilingual speech recognition, speech translation, spoken language identification, and voice activity detection. These tasks are jointly represented as a sequence of tokens to be predicted by the decoder, allowing a single model to replace many stages of a traditional speech-processing pipeline. The multitask training format uses a set of special tokens that serve as task specifiers or classification targets.

Setup

We used Python 3.9.9 and PyTorch 1.10.1 to train and test our models, but the codebase is expected to be compatible with Python 3.8-3.11 and recent PyTorch versions. The codebase also depends on a few Python packages, most notably OpenAI's tiktoken for their fast tokenizer implementation. You can download and install (or update to) the latest release of Whisper with the following command:

pip install -U openai-whisper

Alternatively, the following command will pull and install the latest commit from this repository, along with its Python dependencies:

pip install git+https://github.com/openai/whisper.git 

To update the package to the latest version of this repository, please run:

pip install --upgrade --no-deps --force-reinstall git+https://github.com/openai/whisper.git

It also requires the command-line tool ffmpeg to be installed on your system, which is available from most package managers:

# on Ubuntu or Debian
sudo apt update && sudo apt install ffmpeg

# on Arch Linux
sudo pacman -S ffmpeg

# on MacOS using Homebrew (https://brew.sh/)
brew install ffmpeg

# on Windows using Chocolatey (https://chocolatey.org/)
choco install ffmpeg

# on Windows using Scoop (https://scoop.sh/)
scoop install ffmpeg

You may need rust installed as well, in case tiktoken does not provide a pre-built wheel for your platform. If you see installation errors during the pip install command above, please follow the Getting started page to install Rust development environment. Additionally, you may need to configure the PATH environment variable, e.g. export PATH="$HOME/.cargo/bin:$PATH". If the installation fails with No module named 'setuptools_rust', you need to install setuptools_rust, e.g. by running:

pip install setuptools-rust

Available models and languages

There are five model sizes, four with English-only versions, offering speed and accuracy tradeoffs. Below are the names of the available models and their approximate memory requirements and inference speed relative to the large model; actual speed may vary depending on many factors including the available hardware.

Size Parameters English-only model Multilingual model Required VRAM Relative speed
tiny 39 M tiny.en tiny ~1 GB ~32x
base 74 M base.en base ~1 GB ~16x
small 244 M small.en small ~2 GB ~6x
medium 769 M medium.en medium ~5 GB ~2x
large 1550 M N/A large ~10 GB 1x

The .en models for English-only applications tend to perform better, especially for the tiny.en and base.en models. We observed that the difference becomes less significant for the small.en and medium.en models.

Whisper's performance varies widely depending on the language. The figure below shows a performance breakdown of large-v3 and large-v2 models by language, using WERs (word error rates) or CER (character error rates, shown in Italic) evaluated on the Common Voice 15 and Fleurs datasets. Additional WER/CER metrics corresponding to the other models and datasets can be found in Appendix D.1, D.2, and D.4 of the paper, as well as the BLEU (Bilingual Evaluation Understudy) scores for translation in Appendix D.3.

WER breakdown by language

Command-line usage

The following command will transcribe speech in audio files, using the medium model:

whisper audio.flac audio.mp3 audio.wav --model medium

The default setting (which selects the small model) works well for transcribing English. To transcribe an audio file containing non-English speech, you can specify the language using the --language option:

whisper japanese.wav --language Japanese

Adding --task translate will translate the speech into English:

whisper japanese.wav --language Japanese --task translate

Run the following to view all available options:

whisper --help

See tokenizer.py for the list of all available languages.

Python usage

Transcription can also be performed within Python:

import whisper

model = whisper.load_model("base")
result = model.transcribe("audio.mp3")
print(result["text"])

Internally, the transcribe() method reads the entire file and processes the audio with a sliding 30-second window, performing autoregressive sequence-to-sequence predictions on each window.

Below is an example usage of whisper.detect_language() and whisper.decode() which provide lower-level access to the model.

import whisper

model = whisper.load_model("base")

# load audio and pad/trim it to fit 30 seconds
audio = whisper.load_audio("audio.mp3")
audio = whisper.pad_or_trim(audio)

# make log-Mel spectrogram and move to the same device as the model
mel = whisper.log_mel_spectrogram(audio).to(model.device)

# detect the spoken language
_, probs = model.detect_language(mel)
print(f"Detected language: {max(probs, key=probs.get)}")

# decode the audio
options = whisper.DecodingOptions()
result = whisper.decode(model, mel, options)

# print the recognized text
print(result.text)

More examples

Please use the 🙌 Show and tell category in Discussions for sharing more example usages of Whisper and third-party extensions such as web demos, integrations with other tools, ports for different platforms, etc.

License

Whisper's code and model weights are released under the MIT License. See LICENSE for further details.

More Repositories

1

openai-cookbook

Examples and guides for using the OpenAI API
MDX
52,551
star
2

gym

A toolkit for developing and comparing reinforcement learning algorithms.
Python
33,298
star
3

chatgpt-retrieval-plugin

The ChatGPT Retrieval Plugin lets you easily find personal or work documents by asking questions in natural language.
Python
20,460
star
4

gpt-2

Code for the paper "Language Models are Unsupervised Multitask Learners"
Python
20,335
star
5

CLIP

CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image
Jupyter Notebook
19,524
star
6

openai-python

The official Python library for the OpenAI API
Python
18,300
star
7

gpt-3

GPT-3: Language Models are Few-Shot Learners
15,457
star
8

baselines

OpenAI Baselines: high-quality implementations of reinforcement learning algorithms
Python
15,036
star
9

evals

Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.
Python
13,239
star
10

DALL-E

PyTorch package for the discrete VAE used for DALL·E.
Python
10,428
star
11

shap-e

Generate 3D objects conditioned on text or images
Python
10,285
star
12

triton

Development repository for the Triton language and compiler
C++
9,770
star
13

spinningup

An educational resource to help anyone learn deep reinforcement learning.
Python
8,587
star
14

tiktoken

tiktoken is a fast BPE tokeniser for use with OpenAI's models.
Python
8,533
star
15

universe

Universe: a software platform for measuring and training an AI's general intelligence across the world's supply of games, websites and other applications.
Python
7,385
star
16

jukebox

Code for the paper "Jukebox: A Generative Model for Music"
Python
7,326
star
17

openai-node

The official Node.js / Typescript library for the OpenAI API
TypeScript
6,171
star
18

point-e

Point cloud diffusion for 3D model synthesis
Python
5,777
star
19

consistency_models

Official repo for consistency models.
Python
5,725
star
20

guided-diffusion

Python
5,000
star
21

plugins-quickstart

Get a ChatGPT plugin up and running in under 5 minutes!
Python
4,133
star
22

retro

Retro Games in Gym
C
3,278
star
23

glide-text2im

GLIDE: a diffusion-based text-conditional image synthesis model
Python
3,277
star
24

glow

Code for reproducing results in "Glow: Generative Flow with Invertible 1x1 Convolutions"
Python
3,016
star
25

mujoco-py

MuJoCo is a physics engine for detailed, efficient rigid body simulations with contacts. mujoco-py allows using MuJoCo from Python 3.
Cython
2,586
star
26

openai-quickstart-node

Node.js example app from the OpenAI API quickstart tutorial
JavaScript
2,463
star
27

weak-to-strong

Python
2,252
star
28

improved-gan

Code for the paper "Improved Techniques for Training GANs"
Python
2,218
star
29

improved-diffusion

Release for Improved Denoising Diffusion Probabilistic Models
Python
2,102
star
30

roboschool

DEPRECATED: Open-source software for robot simulation, integrated with OpenAI Gym.
Python
2,064
star
31

image-gpt

Python
1,990
star
32

consistencydecoder

Consistency Distilled Diff VAE
Python
1,933
star
33

finetune-transformer-lm

Code and model for the paper "Improving Language Understanding by Generative Pre-Training"
Python
1,929
star
34

multiagent-particle-envs

Code for a multi-agent particle environment used in the paper "Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments"
Python
1,871
star
35

pixel-cnn

Code for the paper "PixelCNN++: A PixelCNN Implementation with Discretized Logistic Mixture Likelihood and Other Modifications"
Python
1,856
star
36

gpt-2-output-dataset

Dataset of GPT-2 outputs for research in detection, biases, and more
Python
1,828
star
37

human-eval

Code for the paper "Evaluating Large Language Models Trained on Code"
Python
1,755
star
38

requests-for-research

A living collection of deep learning problems
HTML
1,625
star
39

gpt-discord-bot

Example Discord bot written in Python that uses the completions API to have conversations with the `text-davinci-003` model, and the moderations API to filter the messages.
Python
1,569
star
40

openai-quickstart-python

Python example app from the OpenAI API quickstart tutorial
CSS
1,561
star
41

multi-agent-emergence-environments

Environment generation code for the paper "Emergent Tool Use From Multi-Agent Autocurricula"
Python
1,526
star
42

evolution-strategies-starter

Code for the paper "Evolution Strategies as a Scalable Alternative to Reinforcement Learning"
Python
1,504
star
43

generating-reviews-discovering-sentiment

Code for "Learning to Generate Reviews and Discovering Sentiment"
Python
1,491
star
44

neural-mmo

Code for the paper "Neural MMO: A Massively Multiagent Game Environment for Training and Evaluating Intelligent Agents"
Python
1,463
star
45

sparse_attention

Examples of using sparse attention, as in "Generating Long Sequences with Sparse Transformers"
Python
1,347
star
46

maddpg

Code for the MADDPG algorithm from the paper "Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments"
Python
1,284
star
47

prm800k

800,000 step-level correctness labels on LLM solutions to MATH problems
Python
1,154
star
48

Video-Pre-Training

Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos
Python
1,138
star
49

following-instructions-human-feedback

1,098
star
50

universe-starter-agent

A starter agent that can solve a number of universe environments.
Python
1,086
star
51

dalle-2-preview

1,051
star
52

lm-human-preferences

Code for the paper Fine-Tuning Language Models from Human Preferences
Python
1,038
star
53

InfoGAN

Code for reproducing key results in the paper "InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets"
Python
1,029
star
54

procgen

Procgen Benchmark: Procedurally-Generated Game-Like Gym-Environments
C++
959
star
55

supervised-reptile

Code for the paper "On First-Order Meta-Learning Algorithms"
JavaScript
955
star
56

blocksparse

Efficient GPU kernels for block-sparse matrix multiplication and convolution
Cuda
941
star
57

kubernetes-ec2-autoscaler

A batch-optimized scaling manager for Kubernetes
Python
849
star
58

random-network-distillation

Code for the paper "Exploration by Random Network Distillation"
Python
844
star
59

summarize-from-feedback

Code for "Learning to summarize from human feedback"
Python
833
star
60

automated-interpretability

Python
822
star
61

large-scale-curiosity

Code for the paper "Large-Scale Study of Curiosity-Driven Learning"
Python
796
star
62

openai-openapi

OpenAPI specification for the OpenAI API
767
star
63

multiagent-competition

Code for the paper "Emergent Complexity via Multi-agent Competition"
Python
761
star
64

grade-school-math

Python
744
star
65

imitation

Code for the paper "Generative Adversarial Imitation Learning"
Python
643
star
66

deeptype

Code for the paper "DeepType: Multilingual Entity Linking by Neural Type System Evolution"
Python
633
star
67

mlsh

Code for the paper "Meta-Learning Shared Hierarchies"
Python
588
star
68

iaf

Code for reproducing key results in the paper "Improving Variational Inference with Inverse Autoregressive Flow"
Python
499
star
69

mujoco-worldgen

Automatic object XML generation for Mujoco
Python
462
star
70

safety-gym

Tools for accelerating safe exploration research.
Python
421
star
71

vdvae

Repository for the paper "Very Deep VAEs Generalize Autoregressive Models and Can Outperform Them on Images"
Python
407
star
72

coinrun

Code for the paper "Quantifying Transfer in Reinforcement Learning"
C++
372
star
73

robogym

Robotics Gym Environments
Python
362
star
74

weightnorm

Example code for Weight Normalization, from "Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks"
Python
357
star
75

atari-py

A packaged and slightly-modified version of https://github.com/bbitmaster/ale_python_interface
C++
354
star
76

openai-gemm

Open single and half precision gemm implementations
C
335
star
77

vime

Code for the paper "Curiosity-driven Exploration in Deep Reinforcement Learning via Bayesian Neural Networks"
Python
331
star
78

safety-starter-agents

Basic constrained RL agents used in experiments for the "Benchmarking Safe Exploration in Deep Reinforcement Learning" paper.
Python
312
star
79

ebm_code_release

Code for Implicit Generation and Generalization with Energy Based Models
Python
311
star
80

CLIP-featurevis

code for reproducing some of the diagrams in the paper "Multimodal Neurons in Artificial Neural Networks"
Python
294
star
81

gym-http-api

API to access OpenAI Gym from other languages via HTTP
Python
291
star
82

gym-soccer

Python
289
star
83

robosumo

Code for the paper "Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments"
Python
283
star
84

EPG

Code for the paper "Evolved Policy Gradients"
Python
240
star
85

phasic-policy-gradient

Code for the paper "Phasic Policy Gradient"
Python
238
star
86

orrb

Code for the paper "OpenAI Remote Rendering Backend"
C#
235
star
87

miniF2F

Formal to Formal Mathematics Benchmark
Objective-C++
202
star
88

atari-reset

Code for the blog post "Learning Montezuma’s Revenge from a Single Demonstration"
Python
183
star
89

spinningup-workshop

For educational materials related to the spinning up workshops.
TeX
181
star
90

train-procgen

Code for the paper "Leveraging Procedural Generation to Benchmark Reinforcement Learning"
Python
165
star
91

dallify-discord-bot

Example code for using OpenAI’s NodeJS SDK with discord.js SDK to create a Discord Bot that uses Slash Commands.
TypeScript
139
star
92

gym3

Vectorized interface for reinforcement learning environments
Python
136
star
93

lean-gym

Lean
134
star
94

retro-baselines

Publicly releasable baselines for the Retro contest
Python
128
star
95

neural-gpu

Code for the Neural GPU model originally described in "Neural GPUs Learn Algorithms"
Python
120
star
96

baselines-results

Jupyter Notebook
117
star
97

tabulate

public release of Excel / OpenAI API integration
JavaScript
116
star
98

go-vncdriver

Fast VNC driver
Go
116
star
99

human-eval-infilling

Code for the paper "Efficient Training of Language Models to Fill in the Middle"
Python
113
star
100

distribution_augmentation

Code for the paper, "Distribution Augmentation for Generative Modeling", ICML 2020.
Python
108
star