llama VS text-generation-webui

derivative works thereof).”

https://github.com/meta-llama/llama/blob/b8348da38fde8644ef0...

Also even if you did use Llama for something, they could unilaterally pull the rug on you when you got 700 million years, AND anyone who thinks Meta broke their copyright loses their license. (Checking if you are still getting screwed is against the rules)

Therefore, Zuckerberg is accountable for explicitly anticompetitive conduct, I assumed an MMA fighter would appreciate the value of competition, go figure.

One thing I wanted to add and call attention to is the importance of licensing in open models. This is often overlooked when we blindly accept the vague branding of models as “open”, but I am noticing that many open weight models are actually using encumbered proprietary licenses rather than standard open source licenses that are OSI approved (https://opensource.org/licenses). As an example, Databricks’s DBRX model has a proprietary license that forces adherence to their highly restrictive Acceptable Use Policy by referencing a live website hosting their AUP (https://github.com/databricks/dbrx/blob/main/LICENSE), which means as they change their AUP, you may be further restricted in the future. Meta’s Llama is similar (https://github.com/meta-llama/llama/blob/main/LICENSE ). I’m not sure who can depend on these models given this flaw.

It looks like Llama 2 7B took 184,320 A100-80GB GPU-hours to train[1]. This one says it used a 96×H100 GPU cluster for 2 weeks, for 32,256 hours. That's 17.5% of the number of hours, but H100s are faster than A100s [2] and FP16/bfloat16 performance is ~3x better.

If they had tried to replicate Llama 2 identically with their hardware setup, it'd cost a little bit less than twice their MoE model.

[1] https://github.com/meta-llama/llama/blob/main/MODEL_CARD.md#...

Ironically, the LLaMA license text [1] this is lifted verbatim from is itself copyrighted [2] and doesn't grant you the permission to copy it or make changes like s/meta/dbrx/g lol.

[1] https://github.com/meta-llama/llama/blob/main/LICENSE#L65

This is kind of an epistemological debate at this level, and I make an effort to link to some source code [1] any time it seems contentious.

LLMs (of the decoder-only, generative-pretrained family everyone means) are next token predictors in a literal implementation sense (there are some caveats around batching and what not, but none that really matter to the philosophy of the thing).

But, they have some emergent behaviors that are a trickier beast. Probably the best way to think about a typical Instruct-inspired “chat bot” session is of them sampling from a distribution with a KL-style adjacency to the training corpus (sidebar: this is why shops that do and don’t train/tune on MMLU get ranked so differently than e.g. the arena rankings) at a response granularity, the same way a diffuser/U-net/de-noising model samples at the image batch (NCHW/NHWC) level.

The corpus is stocked with everything from sci-fi novels with computers arguing their own sentience to tutorials on how to do a tricky anti-derivative step-by-step.

This mental model has adequate explanatory power for anything a public LLM has ever been shown to do, but that only heavily implies it’s what they’re doing.

There is active research into whether there is more going on that is thus far not conclusive to the satisfaction of an unbiased consensus. I personally think that research will eventually show it’s just sampling, but that’s a prediction not consensus science.

They might be doing more, there is some research that represents circumstantial evidence they are doing more.

[1] https://github.com/meta-llama/llama/blob/54c22c0d63a3f3c9e77...

Predicting subsequent text is pretty much exactly what they do. Lots of very cool engineering that’s a real feat, but at its core it’s argmax(P(token|token,corpus)):

https://github.com/facebookresearch/llama/blob/main/llama/ge...

The engineering feats are up there with anything, but it’s a next token predictor.

https://github.com/facebookresearch/llama/pull/947/

> Instead, it turns out a few hundred lines of Python is genuinely enough to train a basic version!

actually its not just a basic version. Llama 1/2's model.py is 500 lines: https://github.com/facebookresearch/llama/blob/main/llama/mo...

Mistral (is rumored to have) forked llama and is 369 lines: https://github.com/mistralai/mistral-src/blob/main/mistral/m...

and both of these are SOTA open source models.

In transformers, they tried really hard to have a single function or method to deal with both self and cross attention mechanisms, masking, positional and relative encodings, interpolation etc. While it allows a user to use the same function/method for any model, it has led to severe parameter bloat. Just compare the original implementation of llama by FAIR with the implementation by HF to get an idea.

Some of the tools offer a path to doing tool use (fetching URLs and doing things with them) or RAG (searching your documents). I think Oobabooga https://github.com/oobabooga/text-generation-webui offers the latter through plugins.

Our tool, https://github.com/transformerlab/transformerlab-app also supports the latter (document search) using local llms.

You can use webui https://github.com/oobabooga/text-generation-webui

Once you get a version up and running I make a copy before I update it as several times updates have broken my working version and caused headaches.

a decent explanation of parameters outside of reading archive papers: https://github.com/oobabooga/text-generation-webui/wiki/03-%...

a news ai website:

The other answers are recommending paths which give you #1. less control and #2. projects with smaller eco-systems.

If you want a truly general purpose front-end for LLMs, the only good solution right now is oobabooga: https://github.com/oobabooga/text-generation-webui

All other alternatives have only small fractions of the features that oobabooga supports. All other alternatives only support a fraction of the LLM backends that oobabooga supports, etc.

The example waifu in text-generation-webui is good enough for me.

https://github.com/oobabooga/text-generation-webui/blob/main...

> Downloading text-generation-webui takes a minute, let's you use any model and get going.

What you're missing here is you're already in this area deep enough to know what ooogoababagababa text-generation-webui is. Let's back out to the "average Windows desktop user" level. Assuming they even know how to find it:

1) Go to https://github.com/oobabooga/text-generation-webui?tab=readm...

2) See a bunch of instructions opening a terminal window and running random batch/powershell scripts. Powershell, etc will likely prompt you with a scary warning. Then you start wondering who ooobabagagagaba is...

3) Assuming you get this far (many users won't even get to step 1) you're greeted with a web interface[0] FILLED to the brim with technical jargon and extremely overwhelming options just to get a model loaded, which is another mind warp because you get to try to select between a bunch of random models with no clear meaning and non-sensical/joke sounding names from someone called "TheBloke". Ok...

Let's say you somehow braved this gauntlet and get this far now you get to chat with it. Ok, what about my local documents? text-generation-webui itself has nothing for that. Repeat this process over the 10 random open source projects from a bunch of names you've never heard of in an attempt to accomplish that.

This is "I saw this thing from Nvidia explode all over media, twitter, youtube, etc. I downloaded it from Nvidia, double-clicked, pointed it at a folder with documents, and it works".

That's the difference and it's very significant.

[0] - https://raw.githubusercontent.com/oobabooga/screenshots/main...

Maybe a copout answer, but setting up a local LLM on my development machine has been invaluable. I use Deep Seek Coder 6.7 [0] and Oobabooga's UI [1]. It helps me solve simple problems and find bugs, while still leaving the larger architecture decisions to me.

[0] https://huggingface.co/deepseek-ai/deepseek-coder-6.7b-instr...

[1] https://github.com/oobabooga/text-generation-webui

You can download it and run it with [this](https://github.com/oobabooga/text-generation-webui). There's an API mode that you could leverage from your VS Code extension.

Same question here. Ollama is fantastic as it makes it very easy to run models locally, But if you already have a lot of code that processes OpenAI API responses (with retry, streaming, async, caching etc), it would be nice to be able to simply switch the API client to Ollama, without having to have a whole other branch of code that handles Alama API responses. One way to do an easy switch is using the litellm library as a go-between but it’s not ideal (and I also recently found issues with their chat formatting for mistral models).

For an OpenAI compatible API my current favorite method is to spin up models using oobabooga TGW. Your OpenAI API code then works seamlessly by simply switching out the api_base to the ooba endpoint. Regarding chat formatting, even ooba’s Mistral formatting has issues[1] so I am doing my own in Langroid using HuggingFace tokenizer.apply_chat_template [2]

[1] https://github.com/oobabooga/text-generation-webui/issues/53...

[2] https://github.com/langroid/langroid/blob/main/langroid/lang...

Related question - I assume ollama auto detects and applies the right chat formatting template for a model?