[HN Gopher] Show HN: FinetuneDB - AI fine-tuning platform to cre...
___________________________________________________________________
Show HN: FinetuneDB - AI fine-tuning platform to create custom LLMs
Hey HN! We're building FinetuneDB (https://finetunedb.com/), an LLM
fine-tuning platform. It enables teams to easily create and manage
high-quality datasets, and streamlines the entire workflow from
fine-tuning to serving and evaluating models with domain experts.
You can check out our docs here: (https://docs.finetunedb.com/)
FinetuneDB exists because creating and managing high-quality
datasets is a real bottleneck when fine-tuning LLMs. The quality of
your data directly impacts the performance of your fine-tuned
models, and existing tools didn't offer an easy way for teams to
build, organize, and iterate on their datasets. We've been working
closely with our pilot customers, both AI startups and more
traditional businesses like a large newspaper, which is fine-tuning
models on their articles to automate content generation in their
tone of voice. The platform is built with an end-to-end workflow
in mind, from dataset building, fine-tuning, serving, and
evaluating outputs. The centerpiece is a version-controlled, no-
code dataset manager where you can upload existing datasets in
JSONL, use production data, or collaborate with domain experts to
create high-quality datasets for custom use cases. We also offer
evaluation workflows that allow non-technical contributors to
annotate data, review model outputs, and refine responses (LLM-as-
judge also available). We offer: - A free tier for developers and
hobbyists who want to streamline dataset management. - Business-
tier with full feature access for teams, using per-seat pricing. -
A custom tier for model hosting, custom integrations, and self-
hosting. Most users still use OpenAI models, but if you're working
with open-source LLMs, we offer pay-as-you-go pricing for
serverless inference for Llama and Mistral models with up to EUR100
in free credits to get started. We're in public beta right now, so
any feedback--whether it's about features, usability, or anything
else--would be incredibly valuable. If you've worked on fine-
tuning models before or are curious about custom LLMs, we'd love to
hear from you. Our goal is to make the fine-tuning process more
accessible and help more companies leverage their data and domain
experts to create custom LLMs. Thanks for checking it out!
Author : felix089
Score : 44 points
Date : 2024-10-09 15:44 UTC (7 hours ago)
(HTM) web link (finetunedb.com)
(TXT) w3m dump (finetunedb.com)
| farouqaldori wrote:
| Hey all, co-founder here happy to answer any questions!
| namanyayg wrote:
| What benefits does this bring me vs just using OpenAI's
| official tools?
| felix089 wrote:
| Other co-founder here, so we offer more specific features
| around iterating on your datasets and include domain experts
| in this workflow. And I'd argue that you also want your
| datasets not necessarily with your foundation model provider
| like OpenAI, so you have the option to test with and
| potentially switch to open-source models.
| rmbyrro wrote:
| What's the cost of fine tuning and then serving a model, say
| Llama 3 8B or 70B? I couldn't find anything on the website...
| I_am_tiberius wrote:
| Looks nice. What is the price and what does it depend on?
| felix089 wrote:
| Thanks! We have a free tier with limited features. Our pro plan
| starts at EUR50 per seat per month and includes all features.
| Teams often collaborate with domain experts to create datasets.
| And for custom integrations, we offer custom plans on request.
|
| More details here: https://docs.finetunedb.com/getting-
| started/pricing
|
| Any specific features or use cases you're interested in?
| ilovefood wrote:
| Looks pretty cool, congrats so far! Do you allow downloading the
| fine tuned model for local inference?
| felix089 wrote:
| Thank you, and yes that is possible. Which model are you
| looking to fine-tune?
| ilovefood wrote:
| If that's the case then I'll try the platform out :) I want
| to finetune Codestral or Qwen2.5-coder on a custom codebase.
| Thank you for the response! Are there some docs or infos
| about the compatibility of the downloaded models, meaning
| will they work right away with llama.cpp?
| farouqaldori wrote:
| We don't support Codestral or Qwen2.5-coder right out of
| the box for now, but depending on your use-case we
| certainly could add it.
|
| We utilize LoRA for smaller models, and qLoRA (quantized)
| for 70b+ models to improve training speeds, so when
| downloading model weights, what you get is the weights &
| adapter_config.json. Should work with llama.cpp!
| KaoruAoiShiho wrote:
| Am I able to upload a book and have it respond truthfully to the
| book in a way that's superior to NotebookLM or similar? Generally
| most long context solutions are very poor. Or does the data have
| to be in a specific format?
| felix089 wrote:
| To get the outcome you want, RAG (retrieval augmented
| generation) would be the way to go, not fine-tuning. Fine-
| tuning doesn't make the model memorize specific content like a
| book. It teaches new behaviors or styles. RAG allows the model
| to access and reference the book during inference. Our platform
| focuses on fine-tuning with structured datasets, so data needs
| to be in a specific format.
|
| This is a very common topic, so I wrote a blog post that
| explains the difference between fine-tuning and RAG if you're
| interested: https://finetunedb.com/blog/fine-tuning-vs-rag
| thomashop wrote:
| These days, I'd say the easiest and most effective approach
| is to put the whole book in the context of one of the longer
| context models.
| felix089 wrote:
| Agreed, for this use case probably the easiest way to go.
| swyx wrote:
| (and most expensive)
| felix089 wrote:
| Agreed too
| KaoruAoiShiho wrote:
| Not really, for something like gemini the accuracy and
| performance is very poor.
| farouqaldori wrote:
| The magic behind NotebookLM can't be replicated only with
| fine-tuning. It's all about the workflow, from the
| chunking strategy, to retrieval etc.
|
| For a defined specific use-case it's certainly possible
| to beat their performance, but things get harder when you
| try to create a general solution.
|
| To answer your question, the format of the data depends
| entirely on the use-case and how many examples you have.
| The more examples you have, the more flexible you can be.
| cl42 wrote:
| Was looking for a solution like this for a few weeks, and started
| coding my own yesterday. Thank you for launching! Excited to give
| it a shot.
|
| Question: when do you expect to release your Python SDK?
| felix089 wrote:
| Very happy to hear, please do reach out to us with any feedback
| or questions via founders@finetunedb.com
| farouqaldori wrote:
| There hasn't been a significant demand for the Python SDK yet,
| so for now we suggest interacting with the API directly.
|
| With that being said, feel free to email us with your use-case,
| I could build the SDK within a few days!
| cl42 wrote:
| Main requirement is to programmatically send my chat logs.
| Not a big deal though, thanks!
| farouqaldori wrote:
| Ah I see, got it. For now the API should work fine for
| that!
| rmbyrro wrote:
| If you currently have an SDK in any of the 5 major languages,
| or if your API is well documented in a structured way, it
| should be very easy to write ab SDK in Python, Go, anything
| LLMs know well.
| not_alex wrote:
| This looks awesome!
| felix089 wrote:
| Thanks!
___________________________________________________________________
(page generated 2024-10-09 23:00 UTC)