The assertion that LLMs are "capable of surprisingly sophisticated reasoning" is supported with a link an article @willknight wrote on the "Sparks of AGI" paper + criticism of it. >>
LLMS
-
LLMs as Sophisticated Mimics: Knowledge Limits Explored
By
–
And then he describes LLMs this way: "It may be best to think of them as preternaturally knowledgeable and gifted mimics that, although capable of surprisingly sophisticated reasoning, are deeply flawed and have only a limited “knowledge” of the world." >>
-
Llama 2 Fine-tuning: Beyond Code Snippets to Infrastructure
By
–
#Llama2 is bigger, better & ready for use, but don't expect stellar results without #finetuning. There are a lot of posts going around that show fine-tuning in a few lines of #code, but none address the #infra challenge. We do! Tutorial and free trial:
-
Query Your Data Easily with LLMs on Abacus AI Platform
By
–
Here is how easy it is to ask questions and get insights from your data: 1. Connect your data source to our platform
2. We'll process and index your information
3. You can use any of the available LLMs to query it. Check it out https://
abacus.ai -
SHOW-1 and Showrunner Agents: Multi-Agent Simulations for Episodic Content
By
–
To Infinity and Beyond: SHOW-1 and Showrunner Agents in Multi-Agent Simulations
— AK (@_akhaliq) 20 juillet 2023
paper: https://t.co/VKZ8BE7fqy
project page: https://t.co/WGoEDI0iMS
present our approach to generating high-quality episodic content for 2 IP’s (Intellectual Property) using large language models… pic.twitter.com/xzihuj0gUJTo Infinity and Beyond: SHOW-1 and Showrunner Agents in Multi-Agent Simulations paper: https://
fablestudio.github.io/showrunner-age
nts/static/pdfs/To_Infinity_and_Beyond_SHOW-1_And_Showrunner_Agents_in_Multi_Agent_Simulations.pdf
…
project page: https://
fablestudio.github.io/showrunner-age
nts/
… present our approach to generating high-quality episodic content for 2 IP’s (Intellectual Property) using large language models -
Genome-Scale Language Model Predicts Viral Variants
By
–
Please include our genome-scale language model with @arvindr_ @argonne which is open access. Trained on 110 million genomes of viruses and bacteria and able to predict new variants of concern for #SARSCoV2 and other viruses and bacteria. https://
github.com/ramanathanlab/
genslm
… -
Fine-tune Llama 2 models on Replicate platform
By
–
Read more on our blog: https://
replicate.com/blog/fine-tune
-llama-2
… Happy hacking. We can't wait to see what you build. -
Fine-tune Llama 2 on Replicate with cloud GPUs
By
–
You can fine-tune Llama 2 on Replicate. • We actually have GPUs available
• Train in the cloud with one line of code
• Run your trained model with an auto-scaling cloud API https://
replicate.com/blog/fine-tune
-llama-2
… Find out more in the thread below -
Cerebras: 20x Memory Advantage Enables 100B Models Natively
By
–
Because our compute and memory are ~20x larger, even 100B models fit entirely in memory without tensor/pipeline parallel gymnastics. Our GPT implementation uses 1/20th the lines of code of Nvidia Megatron. Huge models work out of the box on Cerebras. This is why G42 picked us.
-
Global Support for Meta’s Open Approach to AI and Llama Platform
By
–
We have a broad range of supporters around the world who believe in our open approach to today’s AI across tech, academia, and policy who see the benefits of Llama and an open platform as we do. See the full statement of support + list of supporters https://
bit.ly/3OkFzNK