News for hackers. Stuff that matters.
Re:Dalai: Automatically install, run, and play with LLaMA on... by EGreg on Sunday March 12, 2023 @08:09PM (#35128598)
Is this 68B of RAM?

How do you get access to that on a Macbook?

Re:Dalai: Automatically install, run, and play with LLaMA on... by junipertea on Sunday March 12, 2023 @08:13PM (#35128663)
That’s 68 billions of parameters. It probably does not fit on ram. Though If you encode each parameter using one byte, you would need 68GB RAM which you could get on workstations at this point.
Re:Dalai: Automatically install, run, and play with LLaMA on... by terafo on Sunday March 12, 2023 @08:22PM (#35128768)
It fits, whisper.cpp uses 4 bit quantization, 13B model takes a little bit more than 8gb and around 9gb ram while inferencing.
Re:Dalai: Automatically install, run, and play with LLaMA on... by taf2 on Sunday March 12, 2023 @08:21PM (#35128765)
Seems to use about 40~ GB RAM here...
Re:Dalai: Automatically install, run, and play with LLaMA on... by gymbeaux on Sunday March 12, 2023 @09:02PM (#35129137)
Everyone with “only” 64GB of RAM is pouting today, including me
Re:Dalai: Automatically install, run, and play with LLaMA on... by Taek on Sunday March 12, 2023 @09:05PM (#35129178)
You can run llama using 4 bits per parameter, 64 GB of RAM is more than enough
Re:Dalai: Automatically install, run, and play with LLaMA on... by geysersam on Monday March 13, 2023 @03:12AM (#35132204)
4 bits is ridiculously little. I'm very curious what makes these models so robust to quantization.
Re:Dalai: Automatically install, run, and play with LLaMA on... by MacsHeadroom on Monday March 13, 2023 @06:56AM (#35133937)
Read The Case for 4 Bit Precision. https://arxiv.org/abs/2212.09720

Spoiler: it's the parameter count. As parameter count goes up, but depth matters less.

It just so happens that at around 10B+ parameters you can quantize down to 4bit with essentially no downsides. Models are that big now. So there's no need to waste RAM by having unnecessary precision for each parameter.

Re:Dalai: Automatically install, run, and play with LLaMA on... by Taek on Monday March 13, 2023 @09:27AM (#35135428)
For completeness, there's also another paper that demonstrated you get more power/accuracy per-bit at 4 bits than at any other level of precision (including 2 bits and 3 bits)
Re:Dalai: Automatically install, run, and play with LLaMA on... by MacsHeadroom on Tuesday March 14, 2023 @12:38AM (#35147506)
That's the paper I referenced. But newer research is already challenging it.

'Int-4 llama is not enough [0] - Int-3 and beyond' suggests 3-bit is best for models larger than ~10B parameters when combining binning and GPTQ.

[0] https://nolanoorg.substack.com/p/int-4-llama-is-not-enough-i...

Re:Dalai: Automatically install, run, and play with LLaMA on... by metadat on Monday March 13, 2023 @12:21AM (#35131018)
What if you have around 400GB of RAM? Would this be enough?
Re:Dalai: Automatically install, run, and play with LLaMA on... by gymbeaux on Monday March 13, 2023 @08:58AM (#35135092)
What I'm referring to requires around 67GB of RAM. With 400GB I would imagine you are in good shape for running most of these GPT-type models.
Re:Dalai: Automatically install, run, and play with LLaMA on... by numpad0 on Monday March 13, 2023 @02:56AM (#35132061)
More like finally "proven right" to have needlessly kept feeding 4/5th of 64GB to Chrome since 2018