Eric Buehler
Eric Buehler
@jett06, my pleasure! I just fixed a small bug (in case you saw the strange behavior), so it should be all ready to go now!
> I'm working on it now.[chenwanqq/candle-llava](https://github.com/chenwanqq/candle-llava) It's not easy dude, tons of image preprocess and tensor concat. Yes! Would you be willing to contribute your implementation here once it's done?
> Yes of course! Great, looking forward to that! I'm working on Idefics 2 here: #309.
Hi @lucasavila00, I'm looking forward to merging this PR! I left one requested change, and there is a conflict which should be pretty easy to resolve. Please let me know...
Can we check if `seq_len>1`? I think that would be pretty reliable.
@lucasavila00 thank you!
@donkey-donkey @staru09 yes exactly! Closing this as there hasn't been any activity for a bit.
@AntBlo memory usage of backprop is very high, what is your GPU memory capacity?
@AntBlo 24GB should be enough for backprop. This may be connected to #21.
Ok. Could I use [this crate](https://docs.rs/sentencepiece/latest/sentencepiece/)? One other question: I am implementing GGUF to HF `tokenizers` conversion in mistral.rs, and have had success with the `unigram` model. I am adding...