Just wanted to come back to this thread with some good news: we’ve listened to your feedback and added logprobs to the Chat Completions endpoint today!
LLMS
-
Chat Completions endpoint now supports logprobs feature
By
–
Just wanted to update this thread with some good news: we’ve added logprobs to the Chat Completions endpoint today! Thanks for the feedback!
-
Chat Completions endpoint now supports logprobs feature
By
–
Thank you for the feedback, we just added logprobs to the Chat Completions endpoint today!
-
Chat Completions API Now Supports Logprobs Feature
By
–
We just added logprobs to Chat Completions today!
-

Mixtral: The Future of Efficient Language Models
By
–
I think the way Mixtral works is likely the future of language models. It seems more efficient and scalable in my opinion.
— Matt Wolfe (@mreflow) 15 décembre 2023
Here's a quick oversimplification of how it works that I clipped from today's news video (on YouTube): pic.twitter.com/zIZXjxY4fkI think the way Mixtral works is likely the future of language models. It seems more efficient and scalable in my opinion. Here's a quick oversimplification of how it works that I clipped from today's news video (on YouTube):
-
Grok AI Tool Name Origin Story and Development
By
–
Great @washingtonpost story about this yesterday by @TaylorLorenz
, who also noted the name Grok was derived from "Grocket." https://
washingtonpost.com/technology/202
3/12/14/grok-grimes-curio-toy/
… -
Apple’s Three-Year Path to Building Its Own Foundation Model
By
–
3 yrs until Apple builds its own foundation model? @munster_gene thinks Apple may try to improve Siri first through an open source model. 3 yrs in this age is like 3 decades. They have to move faster. I assume they are, but we’ll see. They have the data & distribution. https://t.co/ttzJa1qnGs
— Paul Roetzer (@paulroetzer) 15 décembre 20233 yrs until Apple builds its own foundation model? @munster_gene thinks Apple may try to improve Siri first through an open source model. 3 yrs in this age is like 3 decades. They have to move faster. I assume they are, but we’ll see. They have the data & distribution.
-

Fine-tuned Open Source LLMs vs Commercial Models Advantages
By
–
Smaller, faster, fine-tuned #LLMs will soon take over large, #generalAI models. Our latest blog dives into the advantages of #finetuned #opensource LLMs over commercial ones in terms of: Cost Speed & Efficiency Ownership & Transparency https://
pbase.ai/46ZPheC -

Hessian Structure Explains Sharpness Regularization Mysteries
By
–
How Hessian Structure Explains Mysteries in Sharpness Regularization https://
bit.ly/472w1ND
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
What Builders Will Create With Log Probabilities
By
–
Can’t wait to see what y’all build with log probs!!