“Through our MLOps solution, we’ve fine-tuned the process of monitoring machine learning models in real-time. It facilitates a comprehensive view of model performance metrics and how they align with business objectives. If there’s an anomaly in a specific region or sector, it can
MACHINE LEARNING
-
vec2text: Reconstructing Text from Embeddings Paper
By
–
more on this second paper soon but the paper draft is here http://
openreview.net/forum?id=t9dWH
pGkPj
… (hopefully w/ a new version on arxiv later this week) and all the code for both papers is available in vec2text: https://
github.com/jxmorris12/vec
2text
… -
Language Model Inversion Research Talk by Sasha
By
–
Sasha gave this amazing talk on our language model inversion research! 1. text embedding inversion (
http://
arxiv.org/abs/2310.06816)
2. language model output inversion (coming soon…) -
Prepare to switch from OpenAI API to other LLMs
By
–
To all AI teams and individuals who have built services on #OpenAI #API, you must be ready at any moment to switch your service to another service. (Anthropic, LLama, any LLM locally on an instance with guardrails in
-
Building a serverless API infrastructure to call LLMs locally
By
–
It will happen very soon. I need to build a serverless API infrastructure as a service to call LLMs locally on the instances.
-
SegFormer Fine-tuning for Lane Detection in Autonomous Driving
By
–
🚀Blog-tastic Tuesdays🚀https://t.co/PKnXErImfN
— Satya Mallick (@LearnOpenCV) 21 novembre 2023
Lane detection aids with the process of effectively segmenting out lanes from roads, allowing for autonomous driving capabilities. In this research article, dive deeper into the fine-tuning process of HuggingFace SegFormer 🤗using… pic.twitter.com/NmZlyuKKy1Blog-tastic Tuesdays https://
learnopencv.com/segformer-fine
-tuning-for-lane-detection/
… Lane detection aids with the process of effectively segmenting out lanes from roads, allowing for autonomous driving capabilities. In this research article, dive deeper into the fine-tuning process of HuggingFace SegFormer using -
Convolutional Neural Networks Classification Behavior Explained
By
–
Every convnet classifier does that, unless I'm misunderstanding your question.
-
Microsoft’s Orca 2 13B beats 70B models with novel synthetic instruct data approach
By
–
Microsoft’s new 13B Orca 2 outperforms 70B models on many evals. Interesting synthetic instruct data:
1) try many detailed system prompts
2) generate answers using best prompt for each task
3) replace with generic system prompt in tuning demonstrations -
Orca 2: Open-Source Language Models 7B and 13B Parameters
By
–
“Orca 2 comes in two sizes (7 billion and 13 billion parameters); both are created by fine-tuning the corresponding LLAMA 2 base models on tailored, high-quality synthetic data. We open-source Orca 2 to encourage further research…”
-
Skip Connections and Channel Modifications in CV Generative Models
By
–
Pretty much all major CV generative models use skip connections and also modify channel counts over layers