skye0402
skye0402
Same here. 0.1.32 worked, 0.1.33 doesn't. Using `llava:13b-v1.6`. Running on Nvidia T4 (16GB).
Would be definitely a great addition to Ollama: - Concurrency of requests - Using GPU mem for several models I'm running it on cloud using a T4 with 16GB GPU...
Same here - would be great to know the row index of what the user clicked in the original data.
That seems to work, it should be added to the Dockerfile.
Would be great to get this fixed to have a full alternative on ARM64.
I have started to work on this some days ago by cloning their https://github.com/taigaio/taiga-docker I was then using kompose to create the kubernetes files. I then edited/created missing files for...
@MagMueller yes, 100%. Default should be as it's now. For users that want persistence between the sessions they need to specify a folder and state `True`. It behaves then like...
Working on it now. @tgaldes @juanjopc
@MagMueller @gregpr07 I've made the additions and tested it. It's definitely speeding up the whole process as it's cached. I would ignore the `cookies_file `parameter if a browser cache `user_data_dir...
@latslats unfortunately my pull request remains unanswered since weeks. @MagMueller any take on that? It's ok to not use it, but a reply would be appreciated.