#jupyter_notebook#chinese_llm#chinese_nlp#finetune#generative_ai#instruct_gpt#instruction_set#llama#llm#lora#open_models#open_source#open_source_models#qlora
AirLLM is a tool that lets you run very large AI models on computers with limited memory by using a smart layer-by-layer loading technique instead of traditional compression methods. You can run a 70-billion-parameter model on just 4GB of GPU memory, or even a 405-billion-parameter model on 8GB, without losing model quality. The benefit is that you can use powerful AI models on affordable hardware without expensive upgrades, and the tool also offers optional compression features that can speed up performance by up to 3 times while maintaining accuracy.
https://github.com/lyogavin/airllm
🔈Most Common Smart Speaker Usage
People are doing plenty of things with voice commands on their #SmartSpeakers, but shopping isn’t really one of them. Only about a quarter to a third of U.S. smart speaker owners have ever purchased something using voice. Very few do so on a regular basis. An Information article this summer said just 2 percent of #Alexa owners had made a purchase via Alexa in 2018.
🚀@PerspectiveIX via Recode.
❓Do you own a Smart Speaker, like Apple HomePod, Google Home, Amazon Echo, etc?
🗣The State of Chinese Voice Assistants
China is the second-largest consumer market in the world, hitting $4.3T in consumer expenditure in 2016 alone (for reference, the US ranked first with $12.5T, and Japan third with $2.7T).
The Chinese smart home market will reach nearly $23B in 2018, according to Juniper Research. #SmartSpeakers and #AI voice assistants are becoming an integral part of that.
Now that smart speakers are becoming commonplace, the early entrant, #Amazon, is losing its market share — and not just to its rival #Google, but to newer players emerging in China.
🚀@PerspectiveIX via CBInsights.
❓ Do you own a smart speaker?