Qwen2-VL Enhancing Vision-Language Model's Perception of the World at Any Resolution discuss: https://
huggingface.co/papers/2409.12
191
… We present the Qwen2-VL Series, an advanced upgrade of the previous Qwen-VL models that redefines the conventional predetermined-resolution approach in visual
OPEN SOURCE
-
Qwen2-VL: Advanced Vision-Language Model with Dynamic Resolution
By
–
-
HTMX Documentation Quality Raises Open Source Transparency Questions
By
–
More proof that @htmx_org is a CIA psyop. Normal open source projects don't have that kind of documentation, let's be honest.
-
Microsoft Releases GRIN MoE Mixture of Experts Model
By
–
Microsoft releases GRIN MoE GRadient-INformed MoE demo: https://
huggingface.co/spaces/GRIN-Mo
E-Demo/GRIN-MoE
…
model: https://
huggingface.co/microsoft/GRIN
-MoE
…
github: https://
github.com/microsoft/GRIN
-MoE
… With only 6.6B activate parameters, GRIN MoE achieves exceptionally good performance across a diverse set of tasks, particularly in -
SambaNova Cloud Launches Fast API Key for Open Source Models
By
–
Attention, developers! Did you get your fast API key yet? 🔑
— SambaNova (@SambaNovaAI) 18 septembre 2024
Unlock lightning fast inference on the best open source models — all running at full precision on the SambaNova Cloud⚡️🎯
Start developing ⤵️ https://t.co/zm6RCXY00n #AI @aton2006 pic.twitter.com/6xCC3SEco6Attention, developers! Did you get your fast API key yet? Unlock lightning fast inference on the best open source models — all running at full precision on the SambaNova Cloud Start developing http://
cloud.sambanova.ai #AI @aton2006 -
Moshika Female Voice MLX Model Integration Guide
By
–
and add –hf-repo kyutai/moshika-mlx-q to get the moshika (female) voice
-
Solar Pro Preview Now Available for Fine-Tuning on Predibase
By
–
2/ Solar Pro Preview is now available for fine-tuning and serving in Predibase! When fine-tuned, Solar Pro Preview outperforms all other open-source LLMs in our fine-tuning leaderboard.
-
Predibase Introduces Gemma 2, Qwen2, and Phi-3.5 Models
By
–
3/ We’ve introduced a number of models recently including Gemma 2, Qwen2, and Phi-3.5.
-
Predibase Updates: Faster, Easier, Cost-Effective Fine-Tuning and Serving
By
–
1/ Excited to share a host of updates to Predibase that make #finetuning and serving #SLMs #LLMs faster, easier, and more cost-effective than ever! Read more on our blog:
-
LlamaCoder: Open Source App Generator Using Llama 3.1 405B
By
–
With the release of Llama 3.1 405B, @TogetherCompute built LlamaCoder — an open source web app that can generate an entire app from a prompt. The repo has now been cloned by hundreds of devs on GitHub and starred 2K+ times. More on this project ➡️ https://t.co/JBvc6EYoyQ pic.twitter.com/f8YSkJFEvi
— AI at Meta (@AIatMeta) 18 septembre 2024With the release of Llama 3.1 405B, @TogetherCompute built LlamaCoder — an open source web app that can generate an entire app from a prompt. The repo has now been cloned by hundreds of devs on GitHub and starred 2K+ times. More on this project https://
go.fb.me/p5o0x0 -
Hacker News for AI Research Papers Built with o1
By
–
Hacker News of AI research papers built with o1-preview/mini in @Gradio
, added instructions on how to submit a paper