Easiest way to try the 70B without download/install is here at the moment: https://
huggingface.co/spaces/ysharma
/Explore_llamav2_with_TGI
…
@thom_wolf
-
Try Llama 70B Model Online Without Download Installation
By
–
-
Cool AI Image Editing Space on Hugging Face
By
–
really cool space by @linoy_tsaban brb updating my twitter pic https://
huggingface.co/spaces/editing
-images/ledits
… -
vLLM Paged Attention Improves Text Generation Inference Performance
By
–
Though these benchmarks are already out-of-date since vLLM’s paged attention landed in Text-Generation-Inference yesterday https://
github.com/huggingface/te
xt-generation-inference/pull/516
… https://
github.com/huggingface/te
xt-generation-inference/issues/478
… -
LLM Companies Will Proliferate Like Car Manufacturers
By
–
there will soon be as many LLM companies as there are car companies in the world
-
Open-source LLMs extended context length with minimal quality degradation
By
–
What a pleasure to see the open-source and academic community at full speed on pushing smart ways to do long context with pretrained LLM Check this thread and amazing work. Pushing LLaMA up to 8k context and more with negligible degradation in quality
-
SafeTensors Best Practices for Model Storage on HF Hub
By
–
oh I think I know this project 😉 Great to see it being released! don't hesitate to use safetensor to store on the HF hub instead of torch pickle if you can, it's much safer.
-
Collaborative LLM Project with Falcon and LLaMA Teams
By
–
This was a really cool project with @clefourrier @nathanhabib1011 @osanseviero and a lot of help from @slippylolo from the Falcon team and the amazing LLaMA team, in particular Xavier Martinet, Aurélien Rodriguez and Sharan Narang!
-
MMLU Benchmark Updated in Open LLM Leaderboard
By
–
25/ THE END PS: Please, come comment/discuss at https://
huggingface.co/spaces/Hugging
FaceH4/open_llm_leaderboard/discussions/82
… PPS: During the writing of this post, @Ollmer & the community have done an amazing work updating MMLU in the Harness to make it similar to the original. We're thus now updating the leaderboard based on it! -
Open Standardized Benchmarks Essential for AI Model Evaluation
By
–
24/ That's why open, standardized, reproducible benchmarks such as the EleutherAI Harness https://
github.com/EleutherAI/lm-
evaluation-harness/
… or Stanford HELM https://
github.com/stanford-crfm/
helm/
… are invaluable to the community. Without them comparing results across models/papers would be impossible, stifling research! -
MMLU Scores Incomparable: Evaluation Implementation Details Matter
By
–
23/ …not at all comparable even if they're both called MMLU & evaluated on same dataset Takeaway? Evaluations are strongly tied to implementations–down to minute details. A mere indication of "MMLU score" gives almost no information about how you can compare these numbers