Package Details: whisper.cpp-model-medium 1-1

Git Clone URL: https://aur.archlinux.org/whisper.cpp-model-medium.git (read-only, click to copy)
Package Base: whisper.cpp-model-medium
Description: This is an autogenerated file, please see https://github.com/hrehfeld/archlinux-whisper.cpp-model
Upstream URL: https://github.com/ggerganov/whisper.cpp/tree/master/models
Licenses: MIT
Submitter: hrehfeld
Maintainer: hrehfeld
Last Packager: hrehfeld
Votes: 1
Popularity: 0.000099
First Submitted: 2022-12-16 16:54 (UTC)
Last Updated: 2023-11-23 12:44 (UTC)

Dependencies (0)

Required by (0)

Sources (1)

Latest Comments

kearneyBack commented on 2023-12-07 10:07 (UTC) (edited on 2023-12-09 01:09 (UTC) by kearneyBack)

hello, sir. I convert to medium ggml myself and got ggml-medium.bin( fd9727b6e1217c2f614f9b698455c4ffd82463b4), that is same as the web. But if I down from https://ggml.ggerganov.com/ and got ggml-model-whisper-medium-q5_0.bin(7718d4c1ec62ca96998f058114db98236937490e). Both medium model can work. However their file size and sha1sum are diff. issues/1396 tells that "large-q5_0 == quantized large-v2", so I know why the fize size diff. thank you