Santhosh Kumar

If you want to master Python for FREE, Follow this Roadmap:

“If you are beginning to learn Python, here are 5 Books📘 that will put you way ahead of your peers: Thread 🧵 1 of 6

@avikumart_ Yeah truly a beast

@SanthoshKumarS_ Great to see DeciLM is overcoming Mistral-7B in performance. excellent innovation I must say

🌟 Brace yourselves for the revelation of the swiftest and most adept language model - behold, the DeciLM 7B! 🚀 DeciLM 7B, the open-source 7-billion parameter LLM, is changing the speed and accuracy standards by itself. -- Thread --

@freest_man Yeah bro

@SanthoshKumarS_ Deci is awesome 💯

@itsafiz Thanks for checking out @itsafiz

@SanthoshKumarS_ DeciLM is amazing 🚀

@tinztwins Yeah great efficiency as well

@SanthoshKumarS_ DeciLM beats Mistral-7B 🔥🔥

@_mwitiderrick Yup breathtaking

@SanthoshKumarS_ Surprisingly, it does outperform Mistral and Llama

@SanthoshKumarS_ Thanks for sharing!

That's a wrap! & Thank you for Reading If you enjoyed this thread: 1. Follow me @SanthoshKumarS_ for more of this Python & ML Content, 2. RT the tweet below to share this thread with your audience.

Explore it for Yourself! Model Card on Hugging Face : ✔️ huggingface.co/Deci/DeciLM-7B ✔️ huggingface.co/Deci/DeciLM-7B… Demo : ✔️ console.deci.ai/infery-llm-demo Notebook : ✔️ colab.research.google.com/drive/1VU98ezH…

@DataScienceHarp This is Great , Amazed by Low latency and low cost 👏

DeciLM-7B is here and it beats Mistral-7B on all benchmarks - topping the HF Open LLM Leaderboard in it's weight class! Here's what you need to hack around: 📔DeciLM-7B Base Model Notebook: bit.ly/decilm-7b-note… 📘DeciLM-7B Fine-tuning Notebook: bit.ly/decilm-7b-fine… 📗DeciLM-7-Instruct Notebook: bit.ly/declm-7b-instr… Help us get trending on HuggingFace by ❤️ the model cards 🤗 Base model: huggingface.co/Deci/DeciLM-7B 🤗 Instruction-tuned model: huggingface.co/Deci/DeciLM-7B… 🤗 Demo on HF Spaces: huggingface.co/spaces/Deci/De… 👀Check out DeciLM + Infery: hubs.ly/Q02cz_pB0 📑 Read more about the model in the release blog:deci.ai/blog/introduci…

@Sumanth_077 Low latency and low cost make this an excellent combo for high-volume applications. Great Model 👏

Here are all the relevant links: • HuggingFace model: huggingface.co/Deci/DeciLM-7B, huggingface.co/Deci/DeciLM-7B… • Try DeciLM 7B with Infery-LLM demo: console.deci.ai/infery-llm-demo • Starter Notebook: bit.ly/declm-7b-instr…

When DeciLM 7B is combined with Deci's inference SDK (Infery), the throughput improves even higher where it achieves 4.4x greater speed than Mistral 7B with vLLM Checkout the image below: This means - ability to support many users at once while enjoying very low inference cost.

@freest_man Impressed by their Benchmarks 🔥🔥🔥

The Fastest and most Proficient 7-billion parameter base LLM released today Deci has unveiled DeciLM 7B and has set new benchmarks for speed and accuracy. Let's look at some highlights! Unmatched Accuracy: >Scores 61.48 on Open LLM Leaderboard. >Outperforms all current competitors in the 7 billion parameter class. >Surpasses previous leaders Mistral 7B and Llama-2 7B Enhanced Throughput Performance: >Achieves 83% higher throughput than Mistral 7B in PyTorch comparison. >Records 139% increase in throughput over Llama 2 7B. Accelerated Speed with Infery-LLM: >When combined with Infery-LLM, Deci’s Inference SDK, it achieves 4.4x greater speed than Mistral 7B with vLLM >This means - the ability to support many users at once while enjoying very low inference costs. It is beneficial for high-volume customer interactions in telecom, online retail, and cloud services. Innovative Architecture: >Developed using AutoNAC-powered neural architecture search. >Utilizes Variable Grouped Query Attention for optimal accuracy-speed balance. Instruction-Tuned Variant - DeciLM 7B-Instruct: >Achieves 63.09 average on Open LLM Leaderboard. >Instruction-tuned with LoRA on the SlimOrca dataset. Try the new model in the links below 👇

@avikumart_ Deci is Rocking with another great Model

🔊Attention AI enthusiasts🤖🚀 Deci has unveiled DeciLM 7B, a groundbreaking development in the realm of language models. Licensed under Apache 2.0 DeciLM 7B is a fast, accurate and efficient language model out there. Read more👇

@akshay_pachaar Deci Sparkles with yet another great Model 🔥

The fastest and most proficient 7B LLM is here! Introducing DeciLM 7B! 🚀 The new leader in open-source: 7B LLMs. It's redefining benchmarks for speed and accuracy. Here's what makes it special:✨ 🔹Unmatched Accuracy: Achieving an average score of 61.48 on the Open LLM Leaderboard, DeciLM 7B outshines its competitors in the 7-billion parameter class, beating the previous front runner Mistral-7B. 🔸Enhanced Throughput Performance: In a direct PyTorch comparison, DeciLM 7B achieves an 82% increase in throughput over Mistral 7B and a 125% increase in throughput over Llama 2-7B. 🔹DeciLM + Infery LLM = 🚀 DeciLM combined with DeciAI's Inference SDK is a no match for anyone. This powerful duo sets a new standard in throughput performance, achieving speeds 4.4 times greater than Mistral 7B with vLLM This is crucial in sectors like telecommunications, online retail, and cloud services, where real-time response to customer inquiries is vital. Check this out👇 🔸Innovative Architecture: Developed with the assistance of neural architecture search-powered engine, AutoNAC, DeciLM 7B employs Variable Grouped Query Attention, a breakthrough in achieving an optimal balance between accuracy and speed. (Check the image below) 🔹Stellar Instruction-Tuned Variant: DecilLM 7B was instruction-tuned using LoRA on the SlimOrca dataset. The resulting model, DeciLM 7B-Instruct, achieves a remarkable average of 63.09 on the Open LLM Leaderboard. The model is now available on HuggingFace🤗! Find all the details & Starter Notebook in the following tweet! Thanks for reading! 🥂