Structured Outputs with Ollama

[rdescartes]

If anyone needs a more powerful constrain outputs, llama.cpp support gbnf:

https://github.com/ggerganov/llama.cpp/blob/master/grammars/...

[jimmySixDOF]

Thats is exactly what they are using

[lolinder]

Have you found the output for arbitrary grammars to be satisfactory? My naive assumption has been that these models will produce better JSON than other formats simply by virtue of having seen so much of it.

[sa-code]

This is amazing, thank you for the link

[dcreater]

How is it more powerful?

[chirau]

This is wonderful news.

I was actually scratching my head on how to structure a regular prompt to produce csv data without extra nonsense like "Here is your data" and "Please note blah blah" at the beginning and end, so this is much welcome as I can define exactly what I want returned then just push structured output to csv.

[firejake308]

Remember that you still need to include an instruction to produce a CSV to get the prompt into the right context to generate a CSV that makes sense. Otherwise, you may get output that is technically in the CSV format but doesn't make any sense because the model was actually trying to write a paragraph response and the token sampler just selected really low-probability tokens that the model didn't really want to say.

[KTibow]

A lot of the time you can prevent this by prefilling the output with ```\n and stopping at ```.

[quaintdev]

Yay! It works. I used gemma2:2b and gave it below text

   You have spent 190 at Fresh Mart. Current balance: 5098

and it gave below output

   {\n\"amount\": 190,\n\"balance\": 5098 ,\"category\": \"Shopping\",\n\"place\":\"Fresh Mart\"\n}

[diggan]

That's some pretty inconsistent JSON, but I guess that makes sense when using a really small model and gemma on top of that.

[guerrilla]

No way. This is amazing and one of the things I actually wanted. I love ollama be because it makes using an LLM feel like using any other UNIX program. It makes LLMs feel like they belong on UNIX.

Question though. Has anyone had luck running it on AMD GPUs? I've heard it's harder but I really want to support the competition when I get cards next year.

[rcarmo]

Yes, even on iGPUs. I was running it fairly well on a mini-PC with a 780M and the BIOS a set to allocate 16GB of shared memory to it.

[bluechair]

Has anyone seen how these constraints affect the quality of the output out of the LLM?

In some instances, I'd rather parse Markdown or plain text if it means the quality of the output is higher.

[lolinder]

Working with OpenAI's models I've found a very good strategy is to have two passes if you can afford the extra tokens: one pass uses a heavy model and natural language with markdown sections discussing the reasoning and providing a final natural language answer (ideally labeled clearly with a markdown header). The second pass can use a cheaper and faster model to put the answer into a structured output format for consumption by the non-LLM parts of the pipeline.

You basically use JSON schema mode to draw a clean boundary around the wishy-washy language bits, using the LLM as a preprocessor to capture its own output in a useful format.

[mmoskal]

It depends how fine-tuned the model is to JSON output.

Also, you need to tell the model the schema. If you don't you will get more weird tokenization issues.

For example, if the schema expects a JSON key "foobarbaz" and the canonical BPE tokenization is ["foobar", "baz"], the token mask generated by all current constrained output libraries will let the model choose from "f", "foo", "foobar" (assuming these are all valid tokens). The model might then choose "foo", and then the constraint will force eg. "bar" and "baz" as next tokens. Now the model will see ["foo", "bar", "baz"] instead of ["foobar", "baz"] and will get confused [0]

If the model knows from the prompt "foobarbaz" is one of the schema keys, it will generally prefer "foobar" over "foo".

[0] In modern models these tokens are related, because of regularization but they are not the same.

[thot_experiment]

YMMV, it's a negative effect in terms of "reasoning" but the delta isn't super significant in most cases. It really depends on the LLM and whether your prompt is likely to generate a JSON response to begin with, the more you have to coerce the LLM the less likely it is to generate sane input. With smaller models you more quickly end up at the edge of space where the LLM has meaningful predictive power and so the outputs start getting closer to random noise.

FWIW measured by me using a vibes based method, nothing rigorous just a lot of hours spent on various LLM projects. I have not used these particular tools yet but ollama was previously able to guarantee json output through what I assume is similar techniques and my partner and I worked previously on a jsonformer-like thing for oobabooga, another LLM runtime tool.

[parthsareen]

We’ve been keeping a close eye on this as well as research is coming out. We’re looking into improving sampling as a whole on both speed and accuracy.

Hopefully with those changes we might also enable general structure generation not only limited to JSON.

[benreesman]

I can say that I was categorically wrong about the utility of things like instructor.

It’s easy to burn a lot of tokens but if the thing you’re doing merits the cost? You can be a bully with it and while its never the best, 95% as good for zero effort is a tool in one’s kit.

[crystal_revenge]

There was a paper going around claiming that structured outputs did hurt the quality of the output, but it turns out their experiment setup was laughably bad [0].

It looks like, so long as you're reasonable with the prompting, you tend to get better outputs when using structure.

0. https://blog.dottxt.co/say-what-you-mean.html

[nikolayasdf123]

same here. I noticed that when you ask model to generate elaborate responses in natural text, and then come up with an answer, quality is orders of magnitude better, and something in line you would expect human-like reasoning.

asking LLM to directly generate JSON gives much worser results, similar to either random guess or intuition.

[quaintdev]

So I can use this with any supported models? The reason I'm asking is because I can only run 1b-3b models reliably on my hardware.

[parthsareen]

Hey! Author of the blog post here. Yes you should be able to use any model. Your mileage may vary with the smaller models but asking them to “return x in json” tends to help with accuracy (anecdotally).

[JackYoustra]

PRs on this have been open for something like a year! I'm a bit sad about how quiet the maintainers have been on this.

[parthsareen]

Hey! Author of the post and one of the maintainers here. I agree - we (maintainers) got to this late and in general want to encourage more contributions.

Hoping to be more on top of community PRs and get them merged in the coming year.

[dcreater]

Reading tea leaves, they seem to be headed down the corporate path so view everything through that lens and how to maximize profit

[SebastianSosa]

Yeah... I was hoping to collaborate in building this but the conversation reached an abrupt end.

[lxe]

I'm still running oobabooga because of its exlv2 support which does much more efficient inference on dual 3090s

[thot_experiment]

I haven't touched ooba in a while, what's the situation like with exl2 vs the non-homogeneous quantization methods people are using like q3k_s or whatever. IIRC while exl2 is faster the gptq quants were outperforming it in terms of accuracy esp at lower bit depths.

[highlanderNJ]

What's the value-add compared to `outlines`?

https://www.souzatharsis.com/tamingLLMs/notebooks/structured...

[parthsareen]

Hey! Author of the blog here. The current implementation uses llama.cpp GBNF which has allowed for a quick implementation. The biggest value-add at this time was getting the feature out.

With the newer research - outlines/xgrammar coming out, I hope to be able to update the sampling to support more formats, increase accuracy, and improve performance.

[xnx]

Is there a best approach for providing structured input to LLMs? Example: feed in 100 sentences and get each one classified in different ways. It's easy to get structured data out, but my approach of prefixing line numbers seems clumsy.

[mmoskal]

Models are trained on Markdown, JSON and various programming languages, so either one of these should work.

However, in this case, you're best of giving the model sentences one by one to avoid it being confused. If you structure the prompt like "Classify the following sentence, here are the rules ...." + sentence, then you should be hitting prefix cache and get even better performance than when doing a single query. Of course, this only works if you have the prefix cache and are not paying per input token (though most providers now let you indicate you want to use prefix cache and pay less).

[ein0p]

That's very useful. To see why, try to get an LLM _reliably_ generate JSON output without this. Sometimes it will, but sometimes it'll just YOLO and produce something you didn't ask for, that can't be parsed.

[rcarmo]

I must say it is nice to see the curl example first. As much as I like Pydantic, I still prefer to hand-code the schemas, since it makes it easier to move my prototypes to Go (or something else).

[seertaak]

Could someone explain how this is implemented? I saw on Meta's Llama page that the model has intrinsic support for structured output. My 30k ft mental model of LLM is as a text completer, so it's not clear to me how this is accomplished.

Are llama.cpp and ollama leveraging llama's intrinsic structured output capability, or is this something else bolted ex-post on the output? (And if the former, how is the capability guaranteed across other models?)

[evertedsphere]

presumably at each step they mask out all tokens that would be invalid at that step according to the grammar

[vincentpants]

Wow neat! The first step to format ambivalence! Curious to see how well does this perform on the edge, our overhead is always so scarce!

Amazing work as always, looking forward to taking this for a spin!

[lormayna]

This is a fantastic news! I spent hours on fine tuning my prompt to summarise text and output in JSON and still have some issues sometimes. Is this feature available also with Go?

[lioeters]

It looks like the structured output feature is available in Go, with the `format` field.

  type GenerateRequest struct {
    ...
    // Format specifies the format to return a response in.
    Format json.RawMessage `json:"format,omitempty"`

https://github.com/ollama/ollama/blob/de52b6c2f90ff220ed9469...

[diimdeep]

Very annoying marketing and pretending to be anything other than just wrapper around llama.cpp.

[evilduck]

Can you ollama haters stop with this bullshit?

Does llama.cpp do dynamic model loading and unloading? Will it fetch a model you request but isn't downloaded? Does it provide SDKs? Does it have startup services it provides? There's space for things that wrap llama.cpp and solve many of its pain points. You can find piles of reports of people struggling to build and compile llama.cpp for some reason or another who then clicked an Ollama installer and it worked right away.

It's also a free OSS project giving all this away, why are you being an ass and discouraging them?