16 4

MrDragonFox

https://discord.gg/foxengine-ai

AI & ML interests

None yet

Recent Activity

Reacted to danielhanchen's post with 🔥 1 day ago

Vision finetuning is in 🦥Unsloth! You can now finetune Llama 3.2, Qwen2 VL, Pixtral and all Llava variants up to 2x faster and with up to 70% less VRAM usage! Colab to finetune Llama 3.2: https://colab.research.google.com/drive/1j0N4XTY1zXXy7mPAhOC1_gMYZ2F2EBlk?usp=sharing

New activity 21 days ago

m-a-p/MIO-7B-Instruct:code still missing ... at least give us examples

New activity about 1 month ago

LanguageBind/Open-Sora-Plan-v1.3.0:prompt refiner misses part 3 of the model

View all activity

Organizations

MrDragonFox's activity

Reacted to danielhanchen's post with 🔥 1 day ago

Post

1067

1 reply

New activity in m-a-p/MIO-7B-Instruct 21 days ago

code still missing ... at least give us examples

#1 opened 21 days ago by

MrDragonFox

New activity in LanguageBind/Open-Sora-Plan-v1.3.0 about 1 month ago

prompt refiner misses part 3 of the model

#1 opened about 1 month ago by

MrDragonFox

New activity in kyutai/mimi 2 months ago

Training code

#1 opened 2 months ago by

ChristophSchuhmann

New activity in mistral-community/pixtral-12b-240910 2 months ago

Any Inference code?

#6 opened 2 months ago by

DongfuJiang

New activity in G-reen/gpt5o-reflexion-q-agi-llama-3.1-8b 2 months ago

really good model

#2 opened 3 months ago by

gileneo

replied to davanstrien's post 3 months ago

96% to wikipedia - i love the idea - but the similarity estimation is far off

original dataset
https://huggingface.co/datasets/kalomaze/Opus_Instruct_25k

replied to takeraparterer's post 4 months ago

sure and who is gonna pay for that ?

updated a model 5 months ago

SynthoCraft/mini

Updated Jun 29

Reacted to merve's post with 🤗 5 months ago

Post

6004

Fine-tune Florence-2 on any task 🔥

Today we release a notebook and a walkthrough blog on fine-tuning Florence-2 on DocVQA dataset @andito @SkalskiP

Blog: https://huggingface.co/blog 📕
Notebook: https://colab.research.google.com/drive/1hKDrJ5AH_o7I95PtZ9__VlCTNAo1Gjpf?usp=sharing 📖
Florence-2 is a great vision-language model thanks to it's massive dataset and small size!

This model requires conditioning through task prefixes and it's not as generalist, requiring fine-tuning on a new task, such as DocVQA 📝

We have fine-tuned the model on A100 (and one can also use a smaller GPU with smaller batch size) and saw that model picks up new tasks 🥹

See below how it looks like before and after FT 🤩
Play with the demo here andito/Florence-2-DocVQA 🏄‍♀️