Today, we announce our first Large Language Model release of 2024: Stable Code 3B. This new LLM is available for non-commercial & commercial use through our new Stability AI Membership: https://
stability.ai/membership Learn more here: https://
bit.ly/3O4oGWW
LLMS
-
Stability AI Launches Stable Code 3B Large Language Model
By
–
-
12 Best Practices for Distilling Smaller Language Models from GPT
By
–
Ready to graduate from #OpenAI to #opensource? Check out our definitive list of 12 best practices for distilling smaller language models from #GPT.
-
Yi Learning Hub Launch with Community Content Contribution Rewards
By
–
Yi Learning Hub is now live for the Yi community! You can find all the latest Yi model tech content at https://
github.com/01-ai/Yi/blob/
main/docs/learning_hub.md
… Have you created tech content for Yi? If yes, we'd love to see more of your sharing by adding it to Yi Learning Hub. Prizes will go to top contributors! -
LLM Production Challenges: Hallucinations and Agent Error Rates
By
–
We have not seen too many LLM apps and agents in production yet due to LLMs hallucinating or making mistakes. Autonomous agents have way too high of an error rate for any tasks besides straight forward tasks. At http://
Abacus.AI, we have been thinking of this problem; -
LLMs as Backend Services: The New Application Layer
By
–
5. LLMs as Backend, Not Consumer-Facing LLMs should be backend services, not consumer-facing. The prompt engineering required is too onerous for most people. Simple interfaces tailored to tasks unlock the power.Thus the importance of this new application layer.
-
Experimentation Remains Crucial for Model Tuning and Sequencing
By
–
4. Experimentation Still Crucial Dmitry cautions against over relying on current best practices, which change rapidly. Experimentation remains crucial to find the right tuning and sequencing of models for each use case.
-
Tensoic Releases Kan-LLaMA 7B Kannada-Optimized Language Model
By
–
@tensoic team has released Kan-LLaMA [ಕನ್-LLama] — a 7B Llama-2 LoRA PreTrained and FineTuned specifically for "Kannada" tokens. Key highlights: ✦ Llama-2's existing vocabulary has been expanded from 32K tokens to a comprehensive 48k (49420) tokens, enhancing its efficiency
-
Solar LM Newsletter Delivers Daily AI Paper Summaries
By
–
#solarlm #solarinaction delivers paper summaries to your inbox. Solar llm summarizes carefully selected daily papers from @Huggingface Just join at https://
groups.google.com/g/hf-daily-pap
er-newsletter
… thanks to @algo_diver -
Cross-App Memory Persistence and Custom Copilot Development
By
–
The big point will be around the possibility of maintaining the knowledge across conversations in different apps and chats for the same account. It feels similar to an unreleased Personalisation feature on Chat GPT. Besides that, it will also allow you to build custom copilots
-
LLM Hallucinations in Legal Tech: Prevalence and Impact
By
–
Many legal tech startups and law firms now use LLM-based tools for various tasks, yet a core problem remains: hallucinations or the tendency of LLMs to generate content inconsistent with legal facts or principles. How common are they? https://
stanford.io/4aWqqv8