You can imagine a company coming to Facebook, saying it wants a custom version of LLaMA, and Facebook deciding to do the training/fine-tuning/serving for it. All the tech is there, along with the talent, DNA, and resources.
GENERATIVE AI
-
Facebook builds ML-focused Google Cloud competitor with LLaMA
By
–
But there's another part of the story here: Facebook has built what seems to be a very early machine learning-focused prototype of.. Google Cloud. It has a training cluster technology it's developed, hardware for inference, and owns a massively popular OSS-ish LLM, LLaMA.
-
AI Chip Startups Face Setbacks Amid Market Challenges
By
–
Some of the hottest chip startups are now hitting snags. Graphcore laid off 160 people last year as a deal to supply Microsoft fell through, and SambaNova also went through some layoffs earlier this year. Cerebras has pushed hard into LLMs with its own OSS LLM models.
-
DraGAN: Interactive AI-based Image Manipulation
By
–
This is wild:
— AI Breakfast (@AiBreakfast) 19 mai 2023
DraGAN: Interactive point-based manipulation of images using AI.
This gives you controllability of the pose, shape, expression, and layout of the objects in your images. pic.twitter.com/m0Dx08hZ7CThis is wild: DraGAN: Interactive point-based manipulation of images using AI. This gives you controllability of the pose, shape, expression, and layout of the objects in your images.
-
Together Compute and Hugging Face Partnership for Model Inference
By
–
We want to thank our partners @togethercompute for their valuable feedback, for hosting model inference and for providing the @huggingface front-end UI (6/6)
-
BLOOMChat Achieves 66% Preference Over Open-Source LLMs
By
–
1. BLOOMChat achieved a average win rate of 45.25% compared to GPT-4's 54.75% across 6 languages in a human preference study. 2. BLOOMChat was preferred 66% of the time compared to mainstream open-source chat LLMs across 6 languages in a human preference study. (3/6)
-
BLOOMChat Outperforms BLOOM and BLOOMZ in Translation Tasks
By
–
We also evaluated BLOOMChat’s translation capabilities on WMT14-fr-en, WMT14-hi-en and WMT18-zh-en task. BLOOMChat outperforms BLOOM by 31.3% and BLOOMZ by 86.4% (4/6)
-
SambaNova Open Sources BloomChat 176B Multilingual LLM Model
By
–
The model is open source. For more detailed technical report on how we do qualitative evaluation and what we measure, read the blog at https://
sambanova.ai/blog/introduci
ng-bloomchat-176b-the-multilingual-chat-based-llm/
… We are also open sourcing the prompts used to do the qualitative multilingual eval and other code. (5/6) -
BLOOMChat Multilingual Alignment Performance Improvements
By
–
Even though BLOOMChat is trained on English dominated alignment data, we see a significant impact on qualitative metrics for assistant style conversational use cases in multilingual settings. (2/6)
-
BLOOMChat: Open-Source Multilingual Chat LLM Released
By
–
Today we are excited to introduce and open-source BLOOMChat a multilingual chat LLM. Built on top of the BLOOM model (
@BigscienceW
), we further train the model on conversational data from @togethercompute @databricks @laion_ai @huggingface
. Some interesting observations: (1/6)