[HN Gopher] NanoChat - The best ChatGPT that $100 can buy
       ___________________________________________________________________
        
       NanoChat - The best ChatGPT that $100 can buy
        
       https://x.com/karpathy/status/1977755427569111362
        
       Author : huseyinkeles
       Score  : 739 points
       Date   : 2025-10-13 15:22 UTC (7 hours ago)
        
 (HTM) web link (github.com)
 (TXT) w3m dump (github.com)
        
       | daft_pink wrote:
       | Wow, how do we sign up for the Eurekalabs course and how much
       | does it cost?
        
         | huseyinkeles wrote:
         | Karpathy says nanochat will become the capstone project of the
         | course LLM101n being developed by Eureka Labs.
         | 
         | I guess it's still a work in progress? Couldn't find any other
         | information elsewhere.
        
           | Schiphol wrote:
           | A bit more info [here](https://github.com/karpathy/LLM101n)
        
         | karpathy wrote:
         | Still under development, remaining work includes tuning
         | nanochat (current state being solid v0.1) and finalizing the
         | in-between projects so that students can "unlock" all
         | complexity that hides underneath: `torch.Tensor`, `torch.dist`,
         | `.backward()`, '.compile()`, etc. And then the more ops heavy
         | aspects.
        
           | BrokenCogs wrote:
           | What's the pricing for the course/EurekaLabs? P.s. thanks for
           | all you're doing
        
       | karimf wrote:
       | I've always thought about the best way to contribute to humanity:
       | number of people you help x how much you help them. I think what
       | Karpathy is doing is one of the highest leverage ways to achieve
       | that.
       | 
       | Our current world is build on top of open source projects. This
       | is possible because there are a lot of free resources to learn to
       | code so anyone from anywhere in the world can learn and make a
       | great piece of software.
       | 
       | I just hope the same will happen with the AI/LLM wave.
        
         | viccis wrote:
         | I recommend his ANN/LLM from scratch videos to people a lot
         | because not only is he a clear instructor, but his code tends
         | to be very Pythonic and just the right balance of terse but
         | readable (not counting the Pytorch vectorization stuff, but
         | that's not his fault, it's just complex). So I think people
         | benefit just from watching and imitating his code style.
        
         | croes wrote:
         | I'm afraid the technology will do more damage because many
         | people will abuse it for fake news and misinformation.
        
           | IntrepidPig wrote:
           | Yeah it feels similar to inventing the nuke. Or it's even
           | more insidious because the harmful effects of the tech are
           | not nearly as obvious or immediate as the good effects, so
           | less restraint is applied. But also, similar to the nuke,
           | once the knowledge on how to do it is out there, someone's
           | going to use it, which obligates everyone else to use it to
           | keep up.
        
         | shafyy wrote:
         | If it only were so easy
        
         | bkettle wrote:
         | This free tradition in software is I think one of the things
         | that I love so much, but I don't see how it can continue with
         | LLMs due to the extremely high training costs and the powerful
         | hardware required for inference. It just seems like writing
         | software will necessarily require paying rent to the LLM hosts
         | to keep up. I guess it's possible that we'll figure out a way
         | to do local inference in a way that is accessible to everyone
         | in the way that most other modern software tools are, but the
         | high training costs make that seem unlikely to me.
         | 
         | I also worry that as we rely on LLMs more and more, we will
         | stop producing the kind of tutorials and other content aimed at
         | beginners that makes it so easy to pick up programming the
         | manual way.
        
           | hodgesrm wrote:
           | This. It looks like one of the keys to maintaining open
           | source is to ensure OSS developers have access to capable
           | models. In the best of worlds, LLM vendors would recognize
           | that open source software is the commons that feeds their
           | models and ensure it flourishes.
           | 
           | In the real world...
        
           | levocardia wrote:
           | There's a Stephen Boyd quote that's something like "if your
           | optimization problem is too computationally expensive, just
           | go on vacation to Greece for a few weeks and by the time you
           | get back, computers might be fast enough to solve it." With
           | LLMs there's sort of an equivalent situation with cost: how
           | mindblowing would it be able to train this kind of LLM _at
           | all_ even just 4 years ago? And today you can get a
           | kindergartener level chat model for about $100. Not hard to
           | imagine the same model costing $10 of compute in a few years.
           | 
           | There's also a reasonable way to "leapfrog" the training cost
           | with a pre-trained model. So if you were doing nanochat as a
           | learning exercise and had no money, the idea would be to code
           | it up, run one or two very slow gradient descent iterations
           | on your slow machine to make sure it is working, then
           | download a pre-trained version from someone who could spare
           | the compute.
        
             | dingnuts wrote:
             | > today you can get a kindergartener level chat model for
             | about $100. Not hard to imagine the same model costing $10
             | of compute in a few years.
             | 
             | No, it's extremely hard to imagine since I used one of
             | Karpathy's own models to have a basic chat bot like six
             | years ago. Yes, it spoke nonsense; so did my GPT-2 fine
             | tune four years ago and so does this.
             | 
             | And so does ChatGPT
             | 
             | Improvement is linear at best. I still think it's actually
             | a log curve and GPT3 was the peak of the "fun" part of the
             | curve. The only evidence I've seen otherwise is bullshit
             | benchmarks, "agents" that increase performance 2x by
             | increasing token usage 100x, and excited salesmen
             | proclaiming the imminence of AGI
        
               | simonw wrote:
               | Apparently 800 million weekly users are finding ChatGPT
               | useful in its present state.
        
               | infinitezest wrote:
               | 1. According to who? Open AI? 2. Its current state is
               | "basically free and containing no ads". I don't think
               | this will remain true given that, as far as I know, the
               | product is very much not making money.
        
               | simonw wrote:
               | Yes, that number is according to OpenAI. They released
               | that 800m number at DevDay last week.
               | 
               | The most recent leaked annualized revenue rate was
               | $12bn/year. They're spending a lot more than that but
               | convincing customers to hand over $12bn is still a very
               | strong indicator of demand.
               | https://www.theinformation.com/articles/openai-
               | hits-12-billi...
        
           | DennisP wrote:
           | Maybe this isn't possible for LLMs yet, but open source
           | versions of AlphaZero have been trained on peer-to-peer
           | networks.
           | 
           | https://zero.sjeng.org/
           | 
           | https://katagotraining.org/
        
         | martin-t wrote:
         | As noble as the goal sounds, I think it's wrong.
         | 
         | Software is just a tool. Much like a hammer, a knife, or
         | ammonium nitrate, it can be used for both good or bad.
         | 
         | I say this as someone who has spent almost 15 years writing
         | software in my free time and publishing it as open source:
         | building software and allowing anyone to use it does not
         | automatically make other people's lives better.
         | 
         | A lot of my work has been used for bad purposes or what some
         | people would consider bad purposes - cheating on tests,
         | cheating in games, accessing personal information without
         | permission, and in one case my work contributed to someone's
         | doxxing. That's because as soon as you publish it, you lose
         | control over it.
         | 
         | But at least with open source software, every person can use it
         | to the same extent so if the majority of people are good, the
         | result is likely to be more positive than negative.
         | 
         | With what is called AI today, only the largest corporations can
         | afford to train the models which means they are controlled by
         | people who have entirely different incentives from the general
         | working population and many of whom have quite obvious
         | antisocial personality traits.
         | 
         | At least 2 billion people live in dictatorships. AI has the
         | potential to become a tool of mass surveillance and total
         | oppression from which those countries will never recover
         | because just like the models can detect a woman is pregnant
         | before she knows it, it will detect a dissenter long before
         | dissent turns into resistance.
         | 
         | I don't have high hopes for AI to be a force for good and
         | teaching people how toy models work, as fun as it is, is not
         | gonna change it.
        
           | isaacremuant wrote:
           | > At least 2 billion people live in dictatorships. AI has the
           | potential to become a tool of mass surveillance and total
           | oppression from which those countries will never recover
           | because just like the models can detect a woman is pregnant
           | before she knows it, it will detect a dissenter long before
           | dissent turns into resistance.
           | 
           | It already works like this in your precious western
           | democracies and they didn't need AI to be authoritarian total
           | surveillance states in spirit, with quite a lot of support
           | from a propagandized populace that begged for or pretended to
           | agree with the infringement of their civil rights because of
           | terrorism, drugs, covid or protecting the poor poor children.
           | 
           | You can combat tech with legislation and culture but the
           | legislation and culture were way beyond the tech in being
           | extremely authoritian in the first place.
        
           | oliveiracwb wrote:
           | I would genuinely love to think otherwise. But I've seen and
           | grown up seeing good things being used in stupid ways (not
           | necessarily for malice)
        
           | simonw wrote:
           | "With what is called AI today, only the largest corporations
           | can afford to train the models"
           | 
           | I take it you're very positive about Andrej's new project
           | which allows anyone to train a model for a few hundred
           | dollars which is comparable to the state-of-the-art from just
           | 5 years ago then.
        
         | carlcortright wrote:
         | strong +1 - developers like him are heros
        
         | epolanski wrote:
         | Then a single person whose learned those skills decide to
         | poison all of us thanks to the skills acquired.
        
         | contingencies wrote:
         | While documenting a build path is nice, IMHO renting hardware
         | nobody can afford from VC-backed cloud providers using cold
         | hard cash to produce clones of legacy tech using toy datasets
         | under the guise of education is propping up the AI bubble and
         | primarily helping institutional shareholders in those AI bubble
         | companies, particularly their hardware supplier NVidia.
         | Personally I do not see this as helping people or humanity.
         | 
         | This would sit better with me if the repo included a first tier
         | use case for local execution, non-NVidia hardware reference,
         | etc.
        
           | jstummbillig wrote:
           | I think you got your proportions slightly wrong there. This
           | will be contributing as much to an AI bubble as a kid
           | tinkering around with combustion is contribution to global
           | warming.
        
             | contingencies wrote:
             | Not really. Anything that guy does sets the tone for an
             | extended cacophony of fans and followers. It would be a sad
             | day when nobody critically assesses the motivations,
             | effects and framing of those moves. I question the claim
             | this move helps humanity and stand by the assessment it's
             | just more feeding an unfree ecosystem which equates to
             | propping up the bubble.
        
           | simonw wrote:
           | "This would sit better with me if the repo included a first
           | tier use case for local execution, non-NVidia hardware
           | reference, etc."
           | 
           | This is a pretty disheartening way to respond to something
           | like this. Someone puts a great deal of effort into giving
           | something interesting away for free, and is told "you should
           | have also done THIS work for free as well in order for me to
           | value your contribution".
        
             | contingencies wrote:
             | It is an objective and transparent response based on free
             | software world norms. Feel free to interpret differently
             | and to be disheartened. Hell, many of us are disheartened
             | by the AI VC political theater we are seeing right now:
             | experienced programmers, artists, lawyers, perhaps much of
             | humanity. Let's stick to objective elements of the
             | discussion, not emotional opine.
        
           | CamperBob2 wrote:
           | If you can't afford $100 _or_ learn how to train it locally
           | with more time and less money, then this isn 't something you
           | should be focusing on at all.
        
             | contingencies wrote:
             | It is amusing to note the dichotomy between the clearly
             | compassionate, empathetic and altruistic perspective
             | displayed here and the comically overstated framing of
             | helping humanity.
        
               | CamperBob2 wrote:
               | (Shrug) Other sites beckon.
        
         | Yizahi wrote:
         | I would adjust your formula to the:
         | 
         | number of people you help x how much you help them x number of
         | people you harm x how much you harm them
         | 
         | For example - harming a little bit all content creators of the
         | world, by stealing their work without compensation or
         | permission. How much does that cost globally every year after
         | year? How do we even quantify long term consequences of that?
         | Stuff like that.
        
         | Lerc wrote:
         | (This is a bit ranty, but due to a sincere desire for a better
         | world, and being the recipient of personal attacks for
         | believing a better world is achievable by a different path to
         | others)
         | 
         | I feel like this point of view is an ideal not shared by one of
         | the main branches of anti-AI sentiment.
         | 
         | The idea of intellectual property works against this. Rather
         | than contributing to humanity directly, ownership of
         | information is accumulated by individuals and then rented to
         | humanity.
         | 
         | At the same time I agree that people should be able to have a
         | livelihood that affords them the ability to create new
         | intellectual contributions.
         | 
         | The service Karpathy is providing is also being provided by
         | thousands of YouTube creators in a huge variety of topics. It's
         | a little sad that so many must support their efforts with
         | support their efforts with sponsorships from sources with
         | varying degrees of ethical behaviour. Patreon is better but
         | still not ideal. I sincerely believe this _is_ one of the best
         | ways to contribute to society.
         | 
         | A recent Daily Show had Jon Stewart describe training AI as
         | strip mining human knowledge. Training AI is regularly
         | described as theft as if this position is a given without any
         | counter argument possible. It is opinion masquerading as fact.
         | This saddens me because it suggests to me that the war to
         | control the narrative is being won by people who want to
         | entrench a hypercapitalistic vision of ownership where not only
         | is a particular expression of an idea ownable but also stakes a
         | claim to own some of any ideas that come from viewing that
         | expression.
         | 
         | I cannot see any way that this viewpoint would aid humanity as
         | a whole, but instead assign benefits to a collection of
         | individuals. The ability to trade intellectual property means
         | that ownership inevitably gets passed to a smaller and smaller
         | pool of individuals over time.
         | 
         | I think we really do need a new way to consider these issues in
         | light of the modern world. When mentioning these thoughts to
         | others a common refrain is that it doesn't matter because the
         | powers that be (and their lobbyists) will prevent any fix from
         | happening. I have never been fond of that particular fatalism,
         | especially when it inhibits discussion of what would be better.
        
           | oblio wrote:
           | Awesome approach.
           | 
           | I'm all for abolishing IP if all AIs are owned communally.
           | I.e. ideally they're utilities or flat out co-ops like some
           | Spanish businesses.
           | 
           | https://en.wikipedia.org/wiki/Mondragon_Corporation
           | 
           | Consum (Spanish supermarket).
           | 
           | They don't get to use everything communally and then
           | capitalism their way forward.
        
       | flakiness wrote:
       | Eureka Labs: https://github.com/EurekaLabsAI
       | 
       | What a prolific person Andrej is. It's been more than amazing to
       | follow along!
        
       | TheAceOfHearts wrote:
       | Here's the announcement post [0] from Karpathy, which provides a
       | bit of additional context.
       | 
       | [0] https://x.com/karpathy/status/1977755427569111362
        
         | dang wrote:
         | Thanks - we'll put that in the toptext as well
        
       | swyx wrote:
       | > Thank you to chief LLM whisperer Alec Radford for
       | advice/guidance.
       | 
       | oh man an Alec x Andrej podcast would BREAK THE INTERNET... just
       | saying... going from glory days of GPT1 to now building GPT3? in
       | 4 hours
        
         | codybontecou wrote:
         | Please oh please. This would be perfect.
        
       | mhitza wrote:
       | Should be "that you can train for $100"
       | 
       | Curios to try it someday on a set of specialized documents.
       | Though as I understand the cost of running this is whatever GPU
       | you can rent with 80GB of VRAM. Which kind of leaves hobbyists
       | and students out. Unless some cloud is donating gpu compute
       | capacity.
        
         | portaouflop wrote:
         | If I have let's say 40gb RAM does it not work at all or just
         | take twice as long to train?
        
           | typpilol wrote:
           | Won't work at all. Or if it does it'll be so slow since it'll
           | have to go to the disk for every single calculation so it
           | won't ever finish.
        
             | karpathy wrote:
             | It will work great with 40GB GPU, probably a bit less than
             | twice slower. These are micro models of a few B param at
             | most and fit easily during both training and inference.
        
         | Onavo wrote:
         | A GPU with 80GB VRAM costs around $1-3 USD an hour on commodity
         | clouds (i.e. the non-Big 3 bare metal providers e.g.
         | https://getdeploying.com/reference/cloud-gpu/nvidia-h100). I
         | think it's accessible to most middle class users in first world
         | countries.
        
           | antinomicus wrote:
           | Isn't the whole point to run your model locally?
        
             | theptip wrote:
             | No, that's clearly not a goal of this project.
             | 
             | This is a learning tool. If you want a local model you are
             | almost certainly better using something trained on far more
             | compute. (Deepseek, Qwen, etc)
        
             | yorwba wrote:
             | The 80 GB are for training with a batch size of 32 times
             | 2048 tokens each. Since the model has only about 560M
             | parameters, you could probably run it on CPU, if a bit
             | slow.
        
             | jsight wrote:
             | I'd guess that this will output faster than the average
             | reader can read, even while using only CPU inferencing on a
             | modern-ish CPU.
             | 
             | The param count is small enough that even cheap (<$500)
             | GPUs would work too.
        
             | simonw wrote:
             | You can run a model locally on _much_ less expensive
             | hardware. It 's training that requires the really big GPUs.
        
       | Havoc wrote:
       | >If your GPU(s) have less than 80GB, you'll have to tune some of
       | the hyperparameters or you will OOM / run out of VRAM. Look for
       | --device_batch_size in the scripts and reduce it until things
       | fit. E.g. from 32 (default) to 16, 8, 4, 2, or even 1.
       | 
       | That sounds like it could run on a 24gb GPU. Batch size of 8
       | would imply 20gb mem, no?
       | 
       | ...presumably just takes forever
        
         | zipy124 wrote:
         | Yes, you can always stream data when training or doing
         | inference on models when vram is lacking but the slow down is
         | extremely noticeable. This is the case for CPU code too and is
         | why optimising for bandwidth is so critical in high-performance
         | computing. Your ability to compute is almost always
         | substantially larger than your bandwidth. An Avx512 capable CPU
         | with a suitable amount of cores is easily capable of doing
         | multiple terabytes of fp64 operations per second, but is
         | typically limited by memory bandwidth, GPUs with LLMs have just
         | broadened this knowledge to more people.
         | 
         | A fun consequence of the fact that CPUs got faster at a rate
         | quicker than memory is look up tables of pre-computed values
         | used to be common optimisations in code, but now it is almost
         | always quicker to re-compute them than to retrieve a pre-
         | computed value from memory for common use-cases.
        
       | faxmeyourcode wrote:
       | This weekend I just cracked into nanoGPT
       | (https://github.com/karpathy/nanoGPT), an older but fabulous
       | learning exercise where you build and train a crappy shakespeare
       | GPT with ~0.8M parameters on a cpu. Results are about what you'd
       | expect from that, they suck, but you can start to feel the magic,
       | especially if you're not a deep learning professional and you
       | just want to poke around and hack on it.
       | 
       | I started writing up a blog post on my weekend with nanoGPT but
       | it's not done yet... Would have been great to link to here lol oh
       | well
        
         | andrewljohnson wrote:
         | the shakespeare code tuned a little with different training
         | data does a good job of generating Magic The Gathering
         | commander decks
        
           | dmarcos wrote:
           | I like the idea of specific-purpose toy models. How did you
           | tune the code and what dataset you used?
        
           | SeanAnderson wrote:
           | would love more details on this. this is exactly the type of
           | project I'd like to dabble in to get more up to speed.
        
         | ACCount37 wrote:
         | It's a useful exercise. A lot of the good ML work is first
         | validated at small scale.
         | 
         | And this new example goes even further - adds instruction
         | following and tool use SFT, as well as RLVR. Makes for a more
         | useful baseline.
        
       | CountGeek wrote:
       | So could I in practice train it on all my psychology books,
       | materials, reports, case study and research papers and then run
       | it on demand on a 1xH100 node -
       | https://getdeploying.com/reference/cloud-gpu/nvidia-h100 whenever
       | I have a specialised question?
        
         | zipy124 wrote:
         | You could but it would be significantly worse than fine-tuning
         | or RAG with a pre-trained model, or using a smaller model since
         | your dataset would be so small.
        
         | leokeba wrote:
         | You could do that indeed, but the performance would be abysmal.
         | For this kind of use-case, it would be a LOT better to use a
         | small pre-trained model and either fine-tune it on your
         | materials, or use some kind of RAG workflow (possibly both).
        
           | dmix wrote:
           | > it would be a LOT better to use a small pre-trained model
           | and either fine-tune it on your materials, or use some kind
           | of RAG workflow (possibly both).
           | 
           | I noticed NewRelic has a chat feature that does this sort of
           | thing, it's scoped very narrowly down to their website and
           | analytics DSL language, and generates charts/data from their
           | db. I've always wondered how they did that (specifically in
           | terms of set up the training/RAG + guardrails). It's super
           | useful.
        
             | simonw wrote:
             | You might be able to figure that out just by asking it -
             | see if you can get it to spit out a copy of the system
             | prompt or tell you what tools it has access to.
             | 
             | The most likely way of building that would be to equip it
             | with a "search_docs" tool that lets it look up relevant
             | information for your query. No need to train an extra model
             | at all if you do that.
        
         | gojomo wrote:
         | Yes, though it's possible a more-general core model, further
         | enhanced with some other ways to bring those texts-of-interest
         | into the working context, might perform better.
         | 
         | Those other ways to integrate the texts might be some form of
         | RAG or other ideas like Apple's recent 'hierarchical memories'
         | (https://arxiv.org/abs/2510.02375).
        
         | alganet wrote:
         | No.
        
       | cyanydeez wrote:
       | if the AI bubble is anything to be compared to, how is 100$ worth
       | anything in GPT terms.
        
       | computer23 wrote:
       | Has the word ChatGPT become generic? This has nothing to do with
       | OpenAI's ChatGPT.
        
         | simonw wrote:
         | It's a reasonable shortcut for what this project provides:
         | training code, inference code _and_ a ChatGPT-style web
         | interface for chatting with the model.
        
       | dinkblam wrote:
       | from their promotional material:
       | 
       | >> Why is the sky blue? > The sky is blue due to an optical
       | illusion called the Rayleigh Scattering
       | 
       | Rayleigh Scattering is not an illusion but an effect.
       | 
       | > [...] particles are made up of tiny blue and violet particles
       | that cause the light to bend in a particular way.
       | 
       | ugh. no, there are no "tiny blue" particles in the sky.
        
         | simonw wrote:
         | That was the _point_. That example is meant to demonstrate that
         | the model that trained for 4 hours can imitate a conversation
         | but isn 't actually anywhere close to being useful.
        
         | kragen wrote:
         | Where did you find that?
        
           | simonw wrote:
           | It's in this screenshot:
           | https://twitter.com/karpathy/status/1977755430093980034
           | 
           | Edit: direct link to image:
           | https://pbs.twimg.com/media/G3Jjxmba8AA5mSs.jpg
        
             | kragen wrote:
             | Aha, thanks!
        
       | sammyd56 wrote:
       | I'm doing a training run right now (started 20min ago). You can
       | follow it at https://api.wandb.ai/links/sjd333-none/dsv4zkij
       | 
       | Will share the resulting model once ready (4 hours from now) for
       | anyone to test inference.
        
         | royosherove wrote:
         | Cool. Is there a simple "howto" on running this repo with
         | training on W&B for a programmer like me who has never done
         | model training flows? Maybe you could share the steps you took?
        
           | sammyd56 wrote:
           | There's not much to it... it took longer to spin up the cloud
           | machine than it did to kick off the training run. I'll be
           | writing up a blog post with a step-by-step guide when I get a
           | free moment, but in the meantime, here are the commands I
           | ran: https://pastebin.com/sdKVy0NR
        
             | royosherove wrote:
             | Ah I was missing the WANDB_RUN env var. so did not get any
             | logs. thanks!
        
         | Lerc wrote:
         | The comment beside the first chart
         | 
         |  _> Our main measure of progress. Bits per byte is, per
         | Karpathy, "a much better measure than just the typical cross-
         | entropy loss, because it further normalizes the loss on each
         | token by the number of bytes of that token, making the metric
         | tokenizer-invariant"._
         | 
         | Is so blindingly obvious, that I'm ashamed to think that I
         | didn't think do it when trialing my own tokenizer approach on
         | tinystories. I might go back and have a look at how well my
         | tokenizer compared to how well I imagined it compared.
        
           | typpilol wrote:
           | Why hasn't anyone made a tokenizer that's 1 character per
           | token. Is it because it requires an insane amount of compute?
           | 
           | Or would the loss of efficiency make it dumber then modern
           | tokenizers?
        
             | skirmish wrote:
             | Since OpenAI tokenizer is estimated at ~4.2 characters per
             | token, with your proposed "1 char per token tokenizer", the
             | effective context length immediately becomes 4.2 times
             | smaller, and generated output 4.2 times slower (since 4.2
             | times more tokens are needed for the same output). Doesn't
             | look like a good tradeoff.
        
             | SeanAnderson wrote:
             | yes to both.
             | 
             | absolutely requires longer training time and more compute.
             | 
             | once trained, predictions need to hold through many more
             | steps because each step processes one token. if a token
             | early in a sentence heavily implies a token will occur
             | later in the sentence then that awareness needs to be
             | maintained while processing each intermediary token and
             | each step is a bit lossy. the fewer steps you need to take
             | before leveraging that knowledge the better the prediction.
        
           | SeanAnderson wrote:
           | ELI5 for anyone else (I had to have this explained to me):
           | 
           | When you train a language model, it tries to predict the next
           | token.
           | 
           | We measure how good it is at that using loss aka how
           | surprised it was by the real answer.
           | 
           | Different models might use different token lengths. So, if
           | you describe loss relative to tokens then you can't easily
           | compare the performance of two models that use different
           | token lengths.
           | 
           | So, compare loss to bytes of text data instead.
        
       | sbassi wrote:
       | Which data uses for training?
        
         | eranation wrote:
         | I think he mentioned somewhere he used fineweb (I assume this
         | one https://huggingface.co/datasets/HuggingFaceFW/fineweb)
        
         | simonw wrote:
         | karpathy/fineweb-edu-100b-shuffle:
         | https://huggingface.co/datasets/karpathy/fineweb-edu-100b-sh...
         | 
         | Which is derived from HuggingFaceFW/fineweb-edu:
         | https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu
         | 
         | HuggingFaceTB/smol-smoltalk:
         | https://huggingface.co/datasets/HuggingFaceTB/smol-smoltalk
         | 
         | And extra fine-tuning on portions of:
         | 
         | cais/mmlu: https://huggingface.co/datasets/cais/mmlu
         | 
         | openai/gsm8k: https://huggingface.co/datasets/openai/gsm8k
         | 
         | allenai/ai2_arc:
         | https://huggingface.co/datasets/allenai/ai2_arc
        
       | efficax wrote:
       | Try ~300k for an 8xH100 lol
        
       | samus wrote:
       | Andrej Karpathy slays again by spreading knowledge about this
       | important subject to the people!
        
       | lebimas wrote:
       | I see Karpathy, I click
        
       | sieve wrote:
       | Nice! His Shakespeare generator was one of the first projects I
       | tried after ollama. The goal was to understand what LLMs were
       | about.
       | 
       | I have been on an LLM binge this last week or so trying to build
       | a from-scratch training and inference system with two back ends:
       | 
       | - CPU (backed by JAX)
       | 
       | - GPU (backed by wgpu-py). This is critical for me as I am
       | unwilling to deal with the nonsense that is rocm/pytorch. Vulkan
       | works for me. That is what I use with llama-cpp.
       | 
       | I got both back ends working last week, but the GPU back end was
       | buggy. So the week has been about fixing bugs, refactoring the
       | WGSL code, making things more efficient.
       | 
       | I am using LLMs extensively in this process and they have been a
       | revelation. Use a nice refactoring prompt and they are able to
       | fix things one by one resulting in something fully functional and
       | type-checked by astral ty.
        
         | danielmarkbruce wrote:
         | Unwilling to deal with pytorch? You couldn't possibly hobble
         | yourself anymore if you tried.
        
           | sieve wrote:
           | If you want to train/sample large models, then use what the
           | rest of the industry uses.
           | 
           | My use case is different. I want something that I can run
           | quickly on one GPU without worrying about whether it is
           | supported or not.
           | 
           | I am interested in convenience, not in squeezing out the last
           | bit of performance from a card.
        
             | danielmarkbruce wrote:
             | You wildly misunderstand pytorch.
        
               | sieve wrote:
               | What is there to misunderstand? It doesn't even install
               | properly most of the times on my machine. You have to use
               | a specific python version.
               | 
               | I gave up on all tools that depend on it for inference.
               | llama-cpp compiles cleanly on my system for Vulkan. I
               | want the same simplicity to test model training.
        
       | earthnail wrote:
       | This is absolutely fantastic. I really can't wait for the final
       | course to be live. It's in the "shut up and take my money"
       | category. I had so much fun with the nanoGPT videos.
        
       | RobGR wrote:
       | This is an LLM trained using a $100 budget to RENT access to
       | graphics cards. It's not about what you could do BUYING hardware
       | for $100.
        
         | danielmarkbruce wrote:
         | Nowhere does he suggest he is buying hardware.
        
       | montebicyclelo wrote:
       | > nanochat is also inspired by modded-nanoGPT
       | 
       | Nice synergy here, the lineage is: Karpathy's nano-GPT -> Keller
       | Jordan's modded-nanoGPT (a speedrun of training nanoGPT) ->
       | NanoChat
       | 
       | modded-nanoGPT [1] is a great project, well worth checking out,
       | it's all about massively speeding up the training of a small GPT
       | model.
       | 
       | Notably it uses the author's Muon optimizer [2], rather than
       | AdamW, (for the linear layers).
       | 
       | [1] https://github.com/KellerJordan/modded-nanogpt
       | 
       | [2] https://kellerjordan.github.io/posts/muon/
        
         | varunneal wrote:
         | Muon was invented by Keller Jordan (and then optimized by
         | others) for the sake of this speedrunning competition. Even
         | though it was invented less than a year ago, it has already
         | been widely adopted as SOTA for model training
        
           | tbalsam wrote:
           | This is the common belief but not quite correct! The Muon
           | update was proposed by Bernstein as the result of a
           | theoretical paper suggesting concrete realizations of the
           | theory, and Keller implemented it and added practical things
           | to get it to work well (input/output AdamW, aggressive
           | coefficients, post-Nesterov, etc).
           | 
           | Both share equal credit I feel (also, the paper's co-
           | authors!), both put in a lot of hard work for it, though I
           | tend to bring up Bernstein since he tends to be pretty quiet
           | about it himself.
           | 
           | (Source: am experienced speedrunner who's been in these
           | circles for a decent amount of time)
        
           | swyx wrote:
           | sharing some useful resrources for learning Muon (since I'm
           | also just catching up on it)
           | 
           | - https://x.com/leloykun/status/1846842883967692926
           | 
           | - https://www.yacinemahdid.com/p/muon-optimizer-explained-
           | to-a...
        
         | echelon wrote:
         | 8xH100 is pretty wild for a single inference node.
         | 
         | Is this what production frontier LLMs are running inference
         | with, or do they consume even more VRAM/compute?
         | 
         | At ~$8/hr, assuming a request takes 5 seconds to fulfill, you
         | can service roughly 700ish requests. About $0.01 per request.
         | 
         | Is my math wrong?
        
           | vessenes wrote:
           | This is the spec for a training node. The inference requires
           | 80GB of VRAM, so significantly less compute.
        
           | Tepix wrote:
           | As vessenes wrote, that's for training. But a H100 can also
           | process many requests in parallel.
        
       | wyldfire wrote:
       | I would love to take an existing open-weight model and fine-tune
       | it with specific training data along these lines. Can I do that
       | with Qwen or GLM? Is there a ~simple recipe for doing that?
        
       | tdhz77 wrote:
       | These are the time of community posts that are legendary.
        
       | kragen wrote:
       | This is really inspiring! Does anyone have some example of how
       | well or poorly it performs on some example prompts?
        
         | kragen wrote:
         | Simon.incutio.com points out that there are screenshots on
         | https://xcancel.com/karpathy/status/1977755430093980034.
        
       | tehnub wrote:
       | Interesting exchange on the use of AI coding tools:
       | curious how much did you write the code by hand of it?
       | Karpathy: Good question, it's basically entirely hand-written
       | (with tab autocomplete). I tried to use claude/codex agents a few
       | times but they just didn't work well enough at all and net
       | unhelpful, possibly the repo is too far off the data
       | distribution.
       | 
       | https://x.com/karpathy/status/1977758204139331904
        
         | oblio wrote:
         | We're still not ready for ouroboros.
        
         | gyomu wrote:
         | > the repo is too far off the data distribution
         | 
         | ah, this explains why these models have been useless to me this
         | whole time. everything i do is just too far off the data
         | distribution!
        
           | SchemaLoad wrote:
           | Everything is unless your app is a React todolist or leatcode
           | questions.
        
             | SeanAnderson wrote:
             | or a typical CRUD app architecture, or a common design
             | pattern, or unit/integration test scaffolding, or standard
             | CI/CD pipeline definitions, or one-off utility scripts,
             | etc...
             | 
             | Like 80% of writing coding is just being a glorified
             | autocomplete and AI is exceptional at automating those
             | aspects. Yes, there is a lot more to being a developer than
             | writing code, but, in those instances, AI really does make
             | a difference in the amount of time one is able to spend
             | focusing on domain-specific deliverables.
        
         | dude250711 wrote:
         | How convenient! You know, my code is somewhat far off the data
         | distribution too.
        
         | SeanAnderson wrote:
         | This makes sense, right? It's a relatively novel thing to be
         | writing. I don't find it to be a damning remark like other
         | comments here seem to be concluding.
         | 
         | If anything, the fact that Karpathy reached towards
         | Claude/Codex in an attempt to gain value is indicative that, in
         | previous coding efforts, those tools were helpful to him.
        
           | simonw wrote:
           | Yeah, if your goal is "build the tightest 8,000 line
           | implementation of training an LLM from scratch, with a focus
           | on both conciseness and educational value" I don't think it's
           | particularly surprising that Claude/Codex weren't much help.
        
         | rootusrootus wrote:
         | That is a good thing to hear from someone as reputable as
         | Karpathy. The folks who think we're on the cusp of AGI may want
         | to temper their expectations a bit.
         | 
         | I do love Claude Code, because one thing I periodically need to
         | do is write some web code, which is not my favorite type of
         | coding but happens to have incredibly good coverage in the
         | training data. Claude is a much better web developer than I am.
         | 
         | But for digging into the algorithmic core of our automation
         | tooling, it doesn't have nearly as much to work with and makes
         | far more mistakes. Still a net win I'm happy to pay for, even
         | if it's never anything more than my web developer slave.
        
           | bdangubic wrote:
           | > But for digging into the algorithmic core of our automation
           | tooling
           | 
           | What I find fascinating is reading this same thing in other
           | context like "UI guru" will say "I would not let CC touch the
           | UI but I let it rip on algorithmic core of our automation
           | tooling cause it is better at it than me..."
        
         | mrtksn wrote:
         | I still find the chat interface the most effective coding
         | companion honestly, AI integration into IDEs so far have proven
         | unaffected for me.
         | 
         | Xcode also has AI integration, at first looks beautifully done
         | in Apple style, but that too brought me more cognitive load
         | instead of removing some.
         | 
         | Why? Because understanding someone else's code is much more
         | difficult than taking someone else's code and plugging it into
         | yours deliberately. IMHO it creates memories on how you built
         | that thing and increases your understanding when you are hands
         | on. You don't necessarily need to craft the code but have full
         | high level grasp on it.
         | 
         | All the current gen AI coding platforms are the wrong way
         | forward I think. I'm inclined to believe that the AI code
         | shouldn't be read by humans at all, it needs to be like this
         | perfect library you use when you our building something for
         | human consumption.
        
       | dabockster wrote:
       | The title is extremely misleading - you have to rent time on an
       | H100 cluster to get it to work. It is not on-device, and thus not
       | truly $100.
       | 
       | I was really excited, too, until I looked through the readme
       | files and the code.
        
         | simonw wrote:
         | It's about training a model from scratch for $100.
        
         | arkmm wrote:
         | What's misleading about that? You rent $100 of time on an H100
         | to train the model.
        
       | oblio wrote:
       | I wonder, if something like this were trained on Wikipedia, could
       | it become a reliable local Wikipedia search engine, basically?
        
         | simonw wrote:
         | I don't think so. Training on documents is not a great way of
         | building a search engine for those for the information in those
         | documents, because the training process mixes all of that
         | information together in ways that detach the individual words
         | from the source documents they came from.
         | 
         | As usual, if you want an LLM to be able to help search a corpus
         | of text the best way to achieve that is to teach it how to use
         | a search tool against that text.
        
       | zoba wrote:
       | I'm very excited for this. An early question I have: what would
       | need to be done to make this a "thinking" model?
        
       | JKCalhoun wrote:
       | "The fastest way to feel the magic is to run the speedrun script
       | speedrun.sh, which trains and inferences the $100 tier of
       | nanochat. On an 8XH100 node at $24/hr, this gives a total run
       | time of about 4 hours."
       | 
       | I am clueless and don't understand this. Where is the $100 being
       | spent? Some sort of API you have to pay to access? Some sort of
       | virtual hardware you have to rent access to?
        
         | llleeeooo wrote:
         | Renting 8 H100s would cost you about 24/h
        
         | simonw wrote:
         | H100s are expensive NVIDIA GPUs, each costing about $30,000.
         | 8XH100 means you have 8 of those wired together in a big server
         | in a data center somewhere , so around a quarter of a million
         | dollars worth of hardware in a single box.
         | 
         | You need that much hardware because each H100 provides 80GB of
         | GPU-accessible RAM, but to train this model you need to hold a
         | LOT of model weights and training data in memory at once. 80*8
         | = 640GB.
         | 
         | ~$24/hour is how much it costs to rent that machine from
         | various providers.
        
           | calmoo wrote:
           | Perfectly explained, thanks!
        
       | KnowledgeWeaver wrote:
       | Ah, but this is nice project. I'll start hacking once it's easier
       | to fine-tune it with own documents for specific questions. What
       | plaques me, though, is how you prevent the model from answering
       | questions it was not trained for?
        
       | cat_plus_plus wrote:
       | End to end training is a different beast, but finetuning and
       | inference of impressive LLMs like QWEN3 can be done on pretty run
       | of the mill hardware like Apple Silicon macs and gaming PCs if
       | anyone wants a personalized assistant with character. Just ask AI
       | how to finetune AI using unsloth (if using NVIDIA) or MLX (for
       | apple) and it will give you ready to run python scripts.
        
       ___________________________________________________________________
       (page generated 2025-10-13 23:00 UTC)