I forgot to mention the probably most important thing: Finetuning Llama 2 7B this way – only requires 13 GB RAM – and can thus comfortably run on a single GPU
MACHINE LEARNING
-
LoRA Quantization and Evaluation Scripts for Llama-2 Models
By
–
Qualitatively, you can use python generate/lora.py –lora_path '… .pth' –quantize "bnb.nf4" –precision "bf16-true" –checkpoint_dir "…/Llama-2-7b-hf" Quantitatively, you can use the python eval/eval_harness scripts: https://
github.com/Lightning-AI/l
it-gpt/tree/main/eval
… -
Specialists vs Generalists: When Task-Specific Models Outperform GPT
By
–
There are instances where “doing well on one specific task” is better than being “a generalist”. GPT-3.5 and GPT-4 are generalists. They perform reasonably well on many language tasks without having to adapt them to any of those tasks, but their performances are limited in
-
QLoRA Single-GPU Limitations and Multi-GPU Future Plans
By
–
QLoRA right now it's single-GPU. I left multi-GPU support as a future todo since I am currently mostly focused on the NeurIPS LLM challenge. But it might well be working with multi-GPU settings already; it's just that I haven't tested. Could be that it requires switching from
-
Hoffman-Yee Grant Highlights 2026 AI Research Winners
By
–
From understanding AI’s societal impact to developing technologies inspired by human intelligence, the Hoffman-Yee grant supports research aimed at solving some of the most challenging problems in AI. Join us on Sept 19 as we highlight this year’s winners: https://
stanford.io/3QNL6Oi -
Reducing Climate-Related Property Risk Using Artificial Intelligence
By
–
Reducing climate-related property risk with AI
#AI #AIio #BigData #ML #NLU #Futureofwork @DeepLearn007 @albertogaruccio @alvinfoo @andi_staub @andy_lucerne @antgrasso @BernardMarr @enricomolinari http://
ow.ly/PHZe30swOuZ -
Meta’s SeamlessM4T Models Now Available on Hugging Face
By
–
The new SeamlessM4T models from @MetaAI are now available on Hugging Face! https://
huggingface.co/models?search=
facebook/seamless-m4t
… -
Meta Releases SeamlessM4T Translation Model Openly
By
–
We believe SeamlessM4T represents a significant breakthrough and as part of our open approach, today we're publicly releasing this work under a CC BY-NC 4.0 license so that others can continue to build on this important field of study. Get the code https://
bit.ly/3E4mtVZ -
SeamlessM4T: All-in-One Multilingual Multimodal Translation Model
By
–
Introducing SeamlessM4T, the first all-in-one, multilingual multimodal translation model. This single model can perform tasks across speech-to-text, speech-to-speech, text-to-text translation & speech recognition for up to 100 languages depending on the task. Details
-
Fine-tuning and ControlNet Advances for SDXL Text-to-Image AI
By
–
Seeing the fast follow from the community racing to get fine-tuning and ControlNet working for SDXL has been exciting. These techniques bring text-to-image AI to life. Here are some examples of my dog Queso of SD2.1 Dreambooth (left) vs. SDXL LoRA (right). The quality is