phasellm VS DeepSpeed

If you're looking for an open source solution you can customize, check out the ResearchLLM demo: https://phasellm.com/researchllm

Code: https://github.com/wgryc/phasellm/tree/main/demos-and-produc...

Using Claude, Cohere, GPT-4, OpenAssistant. Formally swapping between them using PhaseLLM (open source library similar to LangChain).

Large language model evaluation and workflow framework from Phase AI. (https://github.com/wgryc/phasellm)

Fantastic questions! Re: working/not working at times -- this is still an issue. It's why I'm building PhaseLLM more broadly (https://github.com/wgryc/phasellm) -- need a robust pipeline that can also "reset" parts of itself if an LLM makes errors or mistakes.

You can see my prompts in this file: https://github.com/wgryc/phasellm/blob/main/demos-and-produc... I autogenerate a fairly big starting prompt and keep resubmitting it. It describes the data set extensively, which helps quite a bit.

That being said, a lot more can be done here around prompt optimization + making this more robust.

That being said, you might want to actually run head-to-head tests between models. PhaseLLM (free, open source) allows you to build a workflow and plug and play various models (including Dolly 2.0 and GPT-4). Then you can run tests to see how much worse/better the various LLMs are and if that's acceptable for your use case.

Large language model evaluation and workflow framework from Phase AI. (https://github.com/wgryc/phasellm)

DeepSpeed can handle parallelism concerns, and even offload data/model to RAM, or even NVMe (!?) . I'm surprised I don't see this project used more.

While on the surface, a 192GB Mac Studio seems like a great deal (it's not much more than a 48GB A6000!), there are several reasons why this might not be a good idea:

* I assume most people have never used llama.cpp Metal w/ large models. It will drop to CPU speeds whenever the context window is full: https://github.com/ggerganov/llama.cpp/issues/1730#issuecomm... - while sure this might be fixed in the future, it's been an issue since Metal support was added, and is a significant problem if you are actually trying to actually use it for inferencing. With 192GB of memory, you could probably run larger models w/o quantization, but I've never seen anyone post benchmarks of their experiences. Note that at that point, the limited memory bandwidth will be a big factor.

* If you are planning on using Apple Silicon for ML/training, I'd also be wary. There are multi-year long open bugs in PyTorch[1], and most major LLM libs like deepspeed, bitsandbytes, etc don't have Apple Silicon support[2][3].

You can see similar patterns w/ Stable Diffusion support [4][5] - support lagging by months, lots of problems and poor performance with inference, much less fine tuning. You can apply this to basically any ML application you want (srt, tts, video, etc)

Macs are fine to poke around with, but if you actually plan to do more than run a small LLM and say "neat", especially for a business, recommending a Mac for anyone getting started w/ ML workloads is a bad take. (In general, for anyone getting started, unless you're just burning budget, renting cloud GPU is going to be the best cost/perf, although on-prem/local obviously has other advantages.)

[1] https://github.com/pytorch/pytorch/issues?q=is%3Aissue+is%3A...

[2] https://github.com/microsoft/DeepSpeed/issues/1580

[3] https://github.com/TimDettmers/bitsandbytes/issues/485

[4] https://github.com/AUTOMATIC1111/stable-diffusion-webui/disc...

[5] https://forums.macrumors.com/threads/ai-generated-art-stable...

And https://github.com/microsoft/deepspeed

DeepSpeed Chat: Easy, Fast and Affordable RLHF Training of ChatGPT-like Models at All Scales (https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-chat)