Improved download speed of models

[notnian]

I noticed that the script had trouble downloading the models with a stable and constant speed, so I created a small bash script with wget :)

• Download all models to the default cache
• You can specify a model name as an argument to the script
• There is a resume option set in case of interruption of the download

The urls of the most updated models are here :
https://raw.githubusercontent.com/openai/whisper/main/whisper/__init__.py

#!/bin/bash

whisper_cache="$HOME/.cache/whisper"

declare -A models

models["tiny.en"]="https://openaipublic.azureedge.net/main/whisper/models/d3dd57d32accea0b295c96e26691aa14d8822fac7d9d27d5dc00b4ca2826dd03/tiny.en.pt"
models["tiny"]="https://openaipublic.azureedge.net/main/whisper/models/65147644a518d12f04e32d6f3b26facc3f8dd46e5390956a9424a650c0ce22b9/tiny.pt"
models["base.en"]="https://openaipublic.azureedge.net/main/whisper/models/25a8566e1d0c1e2231d1c762132cd20e0f96a85d16145c3a00adf5d1ac670ead/base.en.pt"
models["base"]="https://openaipublic.azureedge.net/main/whisper/models/ed3a0b6b1c0edf879ad9b11b1af5a0e6ab5db9205f891f668f8b0e6c6326e34e/base.pt"
models["small.en"]="https://openaipublic.azureedge.net/main/whisper/models/f953ad0fd29cacd07d5a9eda5624af0f6bcf2258be67c92b79389873d91e0872/small.en.pt"
models["small"]="https://openaipublic.azureedge.net/main/whisper/models/9ecf779972d90ba49c06d968637d720dd632c55bbf19d441fb42bf17a411e794/small.pt"
models["medium.en"]="https://openaipublic.azureedge.net/main/whisper/models/d7440d1dc186f76616474e0ff0b3b6b879abc9d1a4926b7adfa41db2d497ab4f/medium.en.pt"
models["medium"]="https://openaipublic.azureedge.net/main/whisper/models/345ae4da62f9b3d59415adc60127b97c714f32e89e936602e85993674d08dcb1/medium.pt"
models["large-v1"]="https://openaipublic.azureedge.net/main/whisper/models/e4b87e7e0bf463eb8e6956e646f1e277e901512310def2c24bf0e11bd3c28e9a/large-v1.pt"
models["large-v2"]="https://openaipublic.azureedge.net/main/whisper/models/81f7c96c852ee8fc832187b0132e569d6c3065a3252ed18e56effd0b6a73e524/large-v2.pt"
models["large"]="https://openaipublic.azureedge.net/main/whisper/models/81f7c96c852ee8fc832187b0132e569d6c3065a3252ed18e56effd0b6a73e524/large-v2.pt"

download() {
    wget -c $1 -P $whisper_cache
}

if [ ! -d $whisper_cache ]; then
    mkdir -p $whisper_cache
    echo "$whisper_cache directory created"
else
    echo "$whisper_cache already exists"
fi

if [ -z "$1" ]; then
    for name in ${!models[@]}; do
        download ${models[${name}]}
    done
else
    download ${models[${1}]}
fi

[notnian]

I don't know python well enough to improve/fix the current download