Steve's Place on Nostr: If you run the cuda version of Libre Translate, even though it runs in docker, it ...
If you run the cuda version of Libre Translate, even though it runs in docker, it screams. Very fast responses compared with native. At least, that has been my experience. Caveat: Due to moving mastodon between boxes, the one it's on has an Nvidia 3070ti. The cuda version makes use of the GPU. Huge improvement, as far as I can tell.
#mastodon #LibreTranslate #cuda #Nvidia
Published at
2024-10-29 06:41:14Event JSON
{
"id": "606d2057096b12c28a1e734c9fd44712f3ace0a2fee4e02b33fd8cf4918813ce",
"pubkey": "ee71e016b027b6dbb1b52a48d73fd351ffef106fd40daaca67fb125aa55020b1",
"created_at": 1730184074,
"kind": 1,
"tags": [
[
"t",
"mastodon"
],
[
"t",
"libretranslate"
],
[
"t",
"cuda"
],
[
"t",
"nvidia"
],
[
"proxy",
"https://mastodon.stevesworld.co/users/steter/statuses/113389343518442017",
"activitypub"
]
],
"content": "If you run the cuda version of Libre Translate, even though it runs in docker, it screams. Very fast responses compared with native. At least, that has been my experience. Caveat: Due to moving mastodon between boxes, the one it's on has an Nvidia 3070ti. The cuda version makes use of the GPU. Huge improvement, as far as I can tell.\n\n#mastodon #LibreTranslate #cuda #Nvidia",
"sig": "76815a87f80675150c702e7d7d6cb0d07d62cc76edbec3f88b38c791304b05b6c6a17d1af339873314bd0e45ff57b4e60a4feda8ec47730f9de2bef47c3b8f7c"
}
