[HN Gopher] Run DeepSeek R1 Dynamic 1.58-bit
       ___________________________________________________________________
        
       Run DeepSeek R1 Dynamic 1.58-bit
        
       Author : noch
       Score  : 661 points
       Date   : 2025-01-28 08:52 UTC (14 hours ago)
        
 (HTM) web link (unsloth.ai)
 (TXT) w3m dump (unsloth.ai)
        
       | tarruda wrote:
       | Would be great if the next generation of base models was designed
       | to be inferred with 128GB of VRAM while 8bit quantized (which
       | would fit in the consumer hardware class).
       | 
       | For example, I imagine a strong MoE base with 16 billion active
       | parameters and 6 or 7 experts would keep a good performance while
       | being possible to run on 128GB RAM macbooks.
        
         | Davidzheng wrote:
         | Would be great, but unfortunately i think intelligence at that
         | compute scale will be limit by hardware not its model. Though
         | at hardware limit I would expect it to be roughly human level
         | especially if optimized for a particular domain.
        
           | tarruda wrote:
           | I remember that Llama 3 was trained on data curated by Llama
           | 2 and it resulted in a model with a significant performance
           | boost (even though it was trained by a previous generation
           | model of the same size).
           | 
           | Maybe using a strong reasoning model such as R1 the next
           | generation, even more performance can be extracted from
           | smaller models.
        
             | danielbln wrote:
             | That's already happening, and is in fact even part of the
             | R1 training pipeline. An intermediate small reasoning model
             | churns out training data for RL a larger model, rinse and
             | repeat. Deepseek also showed model distillation with
             | synthetic reasoning data to work quite well.
        
               | alchemist1e9 wrote:
               | It's a pretty neat paradigm and I see an abstract
               | connection to how brains dream and produce their own
               | synthetic training data while sleeping that supplements
               | their real data used while awake.
        
               | Davidzheng wrote:
               | Is your first claim in the R1 paper? I didn't see it when
               | I looked
        
         | danielhanchen wrote:
         | So I remember Deepseek used float8 for training - Character AI
         | also used int8 for training - it is indeed possible, but
         | sometimes training can be unstable - Deepseek to my knowledge
         | is actually the first lab to use float8 at a large scale
         | without causing loss spikes - they used FP8 tensor cores, then
         | every 4th matrix multiply, they accumulated to a FP32
         | accumulator - it seems like the Hopper Tensor Cores
         | accumulation mechanism might not be actual FP32 accumulation. I
         | wrote more here:
         | https://x.com/danielhanchen/status/1872719599029850391
        
       | raghavbali wrote:
       | > Unfortunately if you naively quantize all layers to 1.58bit,
       | you will get infinite repetitions in seed 3407: "Colours with
       | dark Colours with dark Colours with dark Colours with dark
       | Colours with dark" or in seed 3408: "Set up the Pygame's Pygame
       | display with a Pygame's Pygame's Pygame's Pygame's Pygame's
       | Pygame's Pygame's Pygame's Pygame's".
       | 
       | This is really interesting insight (although other works cover
       | this as well). I am particularly amused by the process by which
       | the authors of this blog post arrived at these particular seeds.
       | Good work nonetheless!
        
         | littlestymaar wrote:
         | Can't this kind of repetition be dealt with at the ~~decoder~~
         | (edit: sampler) level, like for any models? (see DRY
         | ~~decoder~~ sampler for instance:
         | https://github.com/oobabooga/text-generation-webui/pull/5677)
        
           | ErikBjare wrote:
           | You can deal with this through various sampling methods, but
           | it doesn't actually fix the fried model.
        
           | danielhanchen wrote:
           | Oh yes one could provide a repetition penalty for example -
           | the issue is it's not just repetition that's the issue. I
           | find it rather forgets what it already saw, and so hence it
           | repeats stuff - it's probably best to backtrack, then delete
           | the last few rows in the KV cache.
           | 
           | Another option is to employ min_p = 0.05 to force the model
           | not to generate low prob tokens - it can help especially in
           | the case when the 1.58bit model generates on average 1/8000
           | tokens or so an "incorrect" token (for eg `score := 0`)
        
           | reichardt wrote:
           | You likely mean sampler, not decoder. And no, the stronger
           | the quantization, the more the output token probabilities
           | diverge from the non-quantized model. With a sampler you
           | can't recover any meaningful accuracy. If you force the
           | sampler to select tokens that won't repeat, you're just
           | trading repetitive gibberish for non-repetitive gibberish.
        
             | littlestymaar wrote:
             | > You likely mean sampler, not decoder.
             | 
             | Indeed, that's posting before being fully awake.
             | 
             | > And no, the stronger the quantization, the more the
             | output token probabilities diverge from the non-quantized
             | model. With a sampler you can't recover any meaningful
             | accuracy.
             | 
             | OF course you can't recover any accuracy, but LLM are in
             | fact prone to this kind of repetition no matter what, this
             | is a known failure mode that's why samplers aimed at
             | avoiding this have been designed over the past few years.
             | 
             | > If you force the sampler to select tokens that won't
             | repeat, you're just trading repetitive gibberish for non-
             | repetitive gibberish.
             | 
             | But it won't necessary be gibberish! even a highly
             | quantized R1 has still much more embedded information than
             | a 14 or even 32B model, so I don't see why it should output
             | more gibberish than smaller models.
        
         | danielhanchen wrote:
         | Hey! :) Coincidentally the seeds I always use are 3407, 3408
         | and 3409 :) 3407 because of https://arxiv.org/abs/2109.08203
         | 
         | I also tried not setting the seeds, but the results are still
         | the same - quantizing all layers seems to make the model forget
         | and repeat everything - I put all examples here:
         | https://docs.unsloth.ai/basics/deepseek-r1-dynamic-1.58-bit#...
        
           | iamnotagenius wrote:
           | would be great to have dynamic quants of V3-non-R1 version,
           | as for some tasks it is good enough. Also would be very
           | interesting to see degradation with dynamic quants on
           | small/medium size MoEs, such as older Deepseek models,
           | Mixtrals, IBM tiny Granite MoE. Would be fun if Granite 1b
           | MoE will still be functioning at 1.58bit.
        
             | danielhanchen wrote:
             | Oh yes multiple people have asked me about this - I'll see
             | what I can do :)
        
       | ThePhysicist wrote:
       | In general, how do you run these big models on cloud hardware? Do
       | you cut them up layer-wise and run slices of layers on individual
       | A100/H100s?
        
         | teekert wrote:
         | Was wondering the same, but for HPC clusters :)
        
         | amelius wrote:
         | You could do that, and add pipelining to improve speed.
        
         | phire wrote:
         | My understanding is with MoE (Mixture of Experts), you can and
         | should shard it horizontally. The whole model is 600GB, but
         | only 37GB is active during the evaluation of any single output
         | token.
         | 
         | So you can load a different active subset of the MoE into each
         | 89GB GPU, sharding it across something like 32 different GPUs
         | (or can you get away with less? Wouldn't be surprised if they
         | can infer on 8x H800 gpus). Some parameters are common, others
         | are independent. Queries can be dynamically routed between
         | GPUs, potentially bouncing between GPUs as much as once per
         | output token, depending on which experts they need to activate.
         | 
         | Though, I suspect it's normal to stick on one MoE subset for
         | several output tokens.
         | 
         | This has a secondary benefit that as long as the routing
         | distribution is random, queries should be roughly load balanced
         | across all GPUs.
        
           | yorwba wrote:
           | Each MoE layer has its own router, and it activates 8 (out of
           | 256) experts at a time. There's no reason to expect all of
           | them to stay on the same GPU, so you're pretty much
           | guaranteed to have to do all-to-all communication between the
           | GPUs in your cluster after every layer for every token.
        
             | phire wrote:
             | Interesting.
             | 
             | I had assumed the performance advantage for MoE came from
             | minimising traffic between GPUs. But if it's per layer
             | routing, then it's going to massively increase inter-gpu
             | traffic compared to vertical slicing.
             | 
             | I guess that means the performance advantage actually comes
             | when batching thousands of queries? The MoE routing would
             | mean that on each MoE layer, each GPU shard gets a batch of
             | queries that will all hit roughly the same subset of
             | experts (and read the same weights from memory). The
             | batches then shuffle between each MoE layer to re-optimise.
             | 
             | It's kind of like GPU raytracing where you get large
             | performance gains by running coherency sorting on rays and
             | batching similar rays together.
        
               | yorwba wrote:
               | The performance advantage comes from doing 1/32 of the
               | floating point operations compared to a dense layer with
               | the same number of parameters.
        
               | iamnotagenius wrote:
               | The performance comes mostly from a fraction of memory
               | bandwidth needed, as LLM are mostly memory constrained.
               | Compute matters too, but usually far less than memory.
        
         | danielhanchen wrote:
         | There are a few ways - the most basic is per layer sharding -
         | DeepSeek uses 3 dense layers, so that can stay on GPU0 (with
         | the embedding layer). There's 58 MoE layers (256 experts, 8
         | activated) and 1 shared expert per layer. GPU1 would house
         | layers 3 to 9, and so on.
         | 
         | Then by using pipeline parallelism, if a new request comes, we
         | simply stick them in a queue - GPUs 0, 1, 2, ..., 8. Request A
         | is at GPU 2, Request B at GPU 1, Request C at GPU 0 and so on.
         | 
         | The other option is tensor parallelism were we split the
         | weights evenly. You could combine pipeline and tensor
         | parallelism as well!
        
       | upghost wrote:
       | Thanks for the run instructions, unsloth. Deepseek is so new it's
       | been breaking most of my builds.
        
         | danielhanchen wrote:
         | Glad they were helpful! :)
        
         | marcodiego wrote:
         | This is an important step. Especially for beginners or people
         | who are not in the loop, being able to easily type some simple
         | commands to download, install dependencies, compile and run
         | everything needed for a LLM AI model gives a feeling sci-fi;
         | it's almost like you can have a helping brain at home.
         | 
         | One thing I've being thinking about doing is to combine one of
         | those LLM models running in llama.cpp, feed it with the output
         | of whisper.cpp and connect its output to some TTS model. I
         | wonder how far from Wheels and Roadie from the Pole Position tv
         | series.
        
       | homarp wrote:
       | see also https://news.ycombinator.com/item?id=42846588
        
       | apples_oranges wrote:
       | Random observation 1: I was running DeepSeek yesterday on my
       | Linux with a RTX 4090 and I noticed that the models should fit
       | into VRAM, which is 24GB. Or they are simply slow. So the Apple
       | shared memory architecture has an advantage here. A 192GB Mx
       | Ultra can load and process large models efficiently.
       | 
       | Random observation 2: It's time to cancel the OpenAI
       | subscription.
        
         | anakaine wrote:
         | I disagree with cancelling the OpenAI subscription. I've been
         | getting some help from o1 for both python and php recently, and
         | o1 was doing massively better for the python stuff (it ran,
         | deepseeks didn't and wont with prompt refinement).
        
           | gradus_ad wrote:
           | Were you running a local model?
        
           | neom wrote:
           | Also for some philosophical stuff DeepSeek just won't do it.
           | I'm working on an essay about spirituality and sometimes it
           | just responds that it doesn't know how to work on those types
           | of problems and we should do something fun like math or
           | games, claud tends to reply with something more like "I have
           | to be honest with you, reincarnation is not real" and ChatGPT
           | doesn't seem to care about that kinda thing at all.
        
             | wqaatwt wrote:
             | Just don't ask it about anything related to Tiananmen
             | square or president Pooh..
             | 
             | I'd guess they didn't quite a bit of fine tuning to censor
             | some more sensitive topics which probably impacts the
             | output quality for other non technical subjects.
        
               | greenavocado wrote:
               | Would fine-tuning by using a LoRA paper over the
               | censorship to a large degree?
        
               | im3w1l wrote:
               | Why even bother decensoring it (except academic curiosity
               | ig)? There are a million other ways you can learn about
               | those subjects.
               | 
               | The people making the model probably don't really give a
               | shit about politics and just did the minimum to avoid
               | being embarassed, but if people start jailbreaking it
               | they will be forced to care.
        
           | mst wrote:
           | IIRC thezvi's summary post on R1 mentioned that R1 is amazing
           | for general reasoning and is very clearly a successful proof
           | of concept/capability but a lot of effort seems to have been
           | put into making o1 Good At Code as a practical matter,
           | whereas R1 seems to have been more a research project which
           | proved out the approaches and then was released without
           | sanding the rough edges off because that wasn't the point.
        
         | danielhanchen wrote:
         | Oh yes 192GB machines should be able these quants (131GB for
         | 1.58bit, 158GB for 1.73bit, 183GB for 2.22bit) well :)
        
           | bradfox2 wrote:
           | Great release Daniel. Applaud the consistency you have shown.
           | 
           | Can you release slightly bigger quant versions? Would enjoy
           | something that runs well on 8x32 v100 and 8x80 A100.
        
             | danielhanchen wrote:
             | Thanks! Oh I did release 4bit quants, 5bit, 6bit etc all at
             | https://huggingface.co/unsloth/DeepSeek-R1-GGUF if that
             | helps - they're not dynamic though but it should function
             | fine :)
        
         | moffkalast wrote:
         | Yes, shared memory is a pretty big leg up since it lets the GPU
         | process the whole model even if the bandwidth is slower which
         | still has some benefits.
         | 
         | Apple's M chips, AMD's Strix Point/Halo chips, Intel's Arc
         | iGPUs, Nvidia's Jetsons. The main issue with all of these
         | though is the lack of raw compute to complement the ability to
         | load insanely large models.
        
         | nodja wrote:
         | While 192GB of ram is appealing, it's also quite expensive at
         | $6000. For that price I rather buy a system with 5 used 3090s,
         | which while being "only" 120GB of VRAM, you benefit from much
         | faster tokens/s and prompt processing speed (the macs are
         | notoriously slow at consuming large contexts).
        
           | miohtama wrote:
           | 5x 3090 is also much more power hungry?
        
             | karamanolev wrote:
             | For personal usage, does it matter though? In most places
             | residential electricity is cheap compared to everything
             | else. In a DC context I feel it matters a lot more compared
             | to the capex.
        
               | thenickdude wrote:
               | 1x 3090 (350W power limit) already makes it feel like I'm
               | running a fan heater under my desk, 5x would be nuts.
        
               | TeMPOraL wrote:
               | Place and time your use right, and you'll save a bit on
               | heating at winter and/or at nights.
        
               | heelix wrote:
               | I think the last time any of my computers had a case was
               | back when I realized the pair of 900gx2 cards I was
               | running was turning my computer into an easy bake.
        
               | ryao wrote:
               | When running inference workloads via something like
               | llama.cpp, only 1 GPU is ever used at a time, so you
               | would have 1 active GPU and 4 idle GPUs. That should make
               | the power usage less insane in practice than you expect.
        
           | orf wrote:
           | Can I use that on the train though? I can with a 128GB
           | MacBook, without it sounding like a helicopter taking off as
           | well.
        
             | zem wrote:
             | you don't need to take ai training quite so literally (:
        
             | jchw wrote:
             | Honestly, if you have a residence of some kind and an
             | Internet connection, you don't need to bring your beefy
             | computer with you everywhere. It is cool to be able to have
             | ridiculously powerful mobile computers, but I don't think I
             | would ever be willing to take a $6,000 laptop anywhere it
             | has a decent chance of being stolen.
        
               | orf wrote:
               | Do you live in a third world country? If so I might
               | agree, but otherwise trains are perfectly safe.
        
               | jchw wrote:
               | I am very happy for you, but laptops get stolen in public
               | in most countries.
        
               | orf wrote:
               | Laptops get stolen on a train? An enclosed, single-
               | direction space that only occasionally allows you to exit
               | between infrequent, long-distance stops? A thing that
               | contains ticket inspectors and a literal guard?
               | 
               | How many laptops have you personally seen be stolen on a
               | train?
        
               | rob_c wrote:
               | People stabbed maybe, but that tends to be more sports
               | related than laptop related. (Yes on a national line(!))
        
               | orf wrote:
               | Yeah, a long, enclosed space with no exits is more
               | amenable to drunken violence than petty theft.
        
               | TeMPOraL wrote:
               | You mean a tight, enclosed, single-direction space,
               | crowded with people who are tired, and/or trying to
               | relax, and/or thinking about the destination, and/or
               | otherwise not particularly focused after hours of travel;
               | a thing that contains ticket inspectors that show up
               | every now and then to _check tickets_ , and from which
               | passengers embark and disembark at dozens point along the
               | length of the thing, simultaneously, with no supervision
               | or security checks.
               | 
               | Depending on the train type and configuration, many
               | actually seem like pickpocket paradise.
        
               | orf wrote:
               | Is your laptop in your pocket?
        
               | afavour wrote:
               | Pickpocketing is a very different proposition. They
               | relying on a lack of awareness, taking your wallet and
               | being long gone before you've even noticed. If someone
               | steals your laptop from in front of you without you even
               | noticing I'd suggest that one is on you.
               | 
               | FWIW I've used my laptop on the train plenty, I've never
               | had anything stolen nor felt in any danger of it.
        
               | TeMPOraL wrote:
               | But would you consider leaving it unattended on your seat
               | and going for lunch to the restaurant car, or for an
               | extended toilet break?
        
               | giantrobot wrote:
               | You might have seen some laptops have screens that fold
               | down, I know MacBooks do. This "clam shell" effect
               | protects the keyboard, trackpad, and even the screen from
               | bumps and jostles. Many laptops when so closed can even
               | fit in a backpack.
               | 
               | So a little trick I figured out is to close my laptop lid
               | and then slide it into a pocket of my backpack. I can
               | then carry it with me when I get up and move around.
               | 
               | So then I can take it with me to eat lunch or an extended
               | toilet break. Maybe some day all laptops will have that
               | feature.
        
               | afavour wrote:
               | ...why would I ever do that? You leave something worth
               | several thousand dollars anywhere in public you're
               | risking losing it. What are we even debating here?
        
               | jchw wrote:
               | Only on Hacker News would I have someone arguing with me
               | that laptop theft is not a concern. You know what, you
               | win. It's your $6,000 laptop, not mine.
        
               | orf wrote:
               | So, zero times then. Ok!
        
               | jchw wrote:
               | For what it's worth, I never once insinuated that a
               | laptop would get stolen on a train, only that I wouldn't
               | want to bring such a laptop into the public in the first
               | place. (Presumably, the laptop doesn't come into and exit
               | existence upon entering and exiting the train, so this
               | remains somewhat of a concern even if trains are
               | involved.)
               | 
               | But yes, you're right. I've never personally seen a
               | laptop get stolen. In fact, most people who _have their
               | laptop get stolen_ never see their laptop get stolen
               | either.
               | 
               | I have, however, had coworkers who've had their laptops
               | stolen. Multiple times.
        
               | bookofjoe wrote:
               | Yes! This goes in my forthcoming blog post "Only on
               | Hacker News..."
               | 
               | Yesterday's entry: "... kind of a mind flex that you
               | noted you used Meta Stories glasses to take that photo."
        
               | criddell wrote:
               | A $6000 laptop doesn't look much different than a $1000
               | laptop. I don't think it's a bigger theft risk than any
               | other laptop.
               | 
               | Make sure the laptop is insured and that full disk
               | encryption is enabled. If it's a Mac, make sure you have
               | it in Find My so you can wipe it remotely if that's
               | something you worry about.
        
               | jchw wrote:
               | Honestly, I didn't bother making a better case for why I
               | wouldn't want a $6,000 laptop in large part because the
               | nerve people have to argue that theft isn't a concern at
               | all made me stubborn. Theft is one reason, but a laptop
               | is also a hell of a lot easier to simply break or lose
               | than a desktop that is permanently installed somewhere,
               | and a desktop is more upgradable and repairable, with
               | typically much more I/O.
               | 
               | Today's _baseline_ laptops are really good as it is.
               | 32-64 GiB of RAM is plenty, and at least on PC laptops
               | you can do it fairly cheaply. Apple has been a consistent
               | year or two ahead in mobile CPU performance but it fell
               | out of my consideration ever since I realized the M1 and
               | 7040 were both very sufficient for any local computation
               | I cared about. (I 'm not going to say I'd specifically go
               | for _less_ efficiency or performance, but it has become
               | significantly lower priority over other things like
               | repairability.)
               | 
               | Not really specifically hating on Apple, here. If I was
               | going to get another Mac it'd be a Mac Mini or Mac Studio
               | probably, ideally with a third-party SSD upgrade to both
               | save on costs and get a slight bit of extra drive
               | performance too. I've definitely considered it, even
               | though I am very far from an Apple fan, just due to the
               | superior value and efficiency they have in many
               | categories.
        
               | crazygringo wrote:
               | Yes, all the time. It's happened to two people I know, in
               | France and in the US.
               | 
               | People get up to use the bathroom or the cafe car, the
               | laptop is left behind for ten minutes, one of the train
               | stops is while they're away from their seat, and someone
               | sees an opportunity, snags it, and gets off at the stop.
               | 
               | This is an actual thing. And if it's worth a thousand
               | bucks then it's very much worth getting off at an earlier
               | stop then you'd planned, and continuing your journey on
               | the next train.
               | 
               | Ticket inspectors or guards are irrelevant. There isn't
               | one in your car 99% of the time.
               | 
               | I don't why you're trying to argue laptop theft on trains
               | in first-world countries isn't a thing. It _absolutely_
               | is.
        
               | vladms wrote:
               | Different regions of the world would see different
               | degrees of responsibilities regarding theft. I would
               | consider absurd to leave unattended in a public space
               | something valuable, considering the effort required to
               | avoid that (that is: taking it with you).
               | 
               | So, yes, theft on trains for people that think they are
               | 100% safe are a thing, but applying the same idea (to
               | assume something is 100% safe and not be cautious) I
               | wonder how do such people use the internet...
        
               | Thorrez wrote:
               | My coworker was having coffee and using his work laptop
               | at an outdoor coffeeshop in Mountain View, CA. Someone on
               | a bike rode by and attempted grab his phone and bike off
               | with it.
               | 
               | The attempted thief didn't succeed in taking the phone,
               | but did knock the laptop onto the ground, damaging it.
        
               | vladms wrote:
               | The discussion was about leaving unattended valuable
               | objects in public places. Sure, a theft can happen even
               | if attended, or using violence, but I personally avoid
               | increasing the chance of having something stolen by
               | leaving it unattended.
               | 
               | If I would make a statistics of primary cause of
               | remaining without a laptop among people I know, the
               | biggest danger is liquids in glasses (that ends up on the
               | laptops) ...
        
               | collingreen wrote:
               | A random person able to dart in and then make a getaway
               | is not what "working on a train" is like and that was the
               | original comment's point.
        
               | crazygringo wrote:
               | You're going to take your laptop with you into the toilet
               | on the train...?
               | 
               | I don't think I've ever seen a human being do that before
               | on a train. Not to go to the toilet, nor to grab a coffee
               | in another car.
               | 
               | You can't be paranoid about everything. My friend in
               | France had put his laptop back into his bag where it
               | wasn't visible and assumed that was good enough, but
               | someone must have seen him do it and just took the whole
               | bag.
               | 
               | You are applying a totally unreasonable standard, to
               | suppose that the thefts were due to unreasonable
               | carelessness. What, do you think someone should take
               | their large luggage into the bathroom too, every time
               | they need to pee?
               | 
               | Talk about victim-blaming.
        
               | vladms wrote:
               | Yes, if I go to the toilet I take my backpack/small bag
               | with me, because usually I have valuable stuff in them
               | and are easy to carry. This does not apply to a large bag
               | (in which I don't put valuable stuff).
               | 
               | The standard is mine and I follow it. The same way I find
               | absurd not to do it, you find it unreasonable to do it.
               | 
               | I find the expectation that things are not stolen (if
               | unsupervised in public places) strange considering the
               | huge amount of inequalities in wealth around even in
               | civilized countries. I do not agree with the idea of
               | stealing, thiefs should be punished, but expecting
               | everybody "to behave" given the situation seems
               | unrealistic to me.
               | 
               | That does not mean that I think that things are stolen
               | 100% of the time. I have a friend that forgot a laptop on
               | a bus (Netherlands) and the driver found it at the end of
               | the line and gave it to lost objects so my friend got it
               | back.
        
               | r00fus wrote:
               | Ok - that's really poor opsec. If I'm going to the
               | bathroom in a train with my laptop (whether it's
               | expensive or not - it has access to all my stuff - which
               | is arguably more valuable), I'll sleep it, put it in my
               | backpack and take the backpack to the bathroom with me.
               | 
               | My work policies state you simply cannot leave your
               | laptop out of sight for any period unless it's in a
               | secure location (work|home). I feel the same way for my
               | personal laptop as well.
        
               | michaelt wrote:
               | You don't hear much about laptop thefts these days
               | because phones are more valuable, more numerous, and much
               | easier to steal.
               | 
               | Obviously, nobody steals things while the train is in
               | motion. They wait until the train is about to leave the
               | station, snatch a phone or handbag and jump out just as
               | the door is closing. The train leaves, the thief blends
               | in with other passenger leaving the station, and by the
               | time news of the theft has made it from the passengers to
               | the driver to the station staff the thief is long gone.
               | 
               | Of course people drive around $6,000+ cars all the time,
               | so....
        
               | FranchuFranchu wrote:
               | > Obviously, nobody steals things while the train is in
               | motion.
               | 
               | Something interesting: I live near a train line where the
               | doors are not automatic (they have to be opened manually
               | on each stop), and there have been incidents where people
               | get pickpocketed while the train is still in motion, and
               | the thief jumps out right before the station, when the
               | train has slowed down significantly but is still in
               | motion. Many people have been hurt doing this.
        
               | jillyboel wrote:
               | it must be amazing to have so much faith in people like
               | you seem to have
        
               | bongodongobob wrote:
               | For real? Grab it before the door closes.
        
             | immibis wrote:
             | You can use a desktop computer on a train if it's one with
             | power outlets. Might get some funny looks, but I've seen it
             | happen (or at least pictures). :)
        
               | rob_c wrote:
               | Only time I've seen that done was with assistive tech and
               | I do sympathise that those setups are difficult enough
               | with desktops
        
             | diggan wrote:
             | > Can I use that on the train though? I can with a 128GB
             | MacBook, without it sounding like a helicopter taking off
             | as well.
             | 
             | What kind of timescale do you expect to be able to train a
             | useful LLM with that?
        
               | orf wrote:
               | Well it's about an hour to commute on the train so I
               | guess that long :3
        
             | wqaatwt wrote:
             | If you have an internet connection then sure you can?
        
           | danielhanchen wrote:
           | The good thing is since MoEs are mainly memory bound, we just
           | need (VRAM + RAM) to be in the range of 80GB or so in my
           | tests for at least 5 tokens or so /s.
           | 
           | It's better to get (VRAM + RAM) >= 140GB for at least 30 to
           | 40 tokens/s, and if VRAM >= 140GB, then it can approach 140
           | tokens/s!
           | 
           | Another trick is to accept more than 8 experts per pass -
           | it'll be slower, but might be more accurate. You could even
           | try reducing the # of experts to say 6 or 7 for low FLOP
           | machines!
        
           | yobid20 wrote:
           | The power requirement for 5x5090s is 10x higher , so you'll
           | spend far more than $6000 in electricity over time.
        
           | xbmcuser wrote:
           | I think just getting nvidia Project Digits might be the best
           | option. A lot of people when it was announced were
           | underwhelmed. But I think now it could be just the thing for
           | people making their own ai home servers.
           | 
           | https://www.nvidia.com/en-us/project-digits/
        
             | sourcecodeplz wrote:
             | Yep, I think the same. With 128GB fast memory one could run
             | this.
        
           | phkahler wrote:
           | >> While 192GB of ram is appealing, it's also quite expensive
           | at $6000.
           | 
           | That's because it's Apple. It time to start moving to AMD
           | systems with shared memory. My Zen 3 APU system has 64GB
           | these days and its a mini ITX board.
        
             | Maakuth wrote:
             | What is the performance in ML workloads like on AMD APUs
             | compared to Apple Silicon?
        
         | ant6n wrote:
         | So I'm thinking, inference seems mostly memory bound. With a
         | fast CPU (for example 7950x with 16 cores), and 256GB of RAM
         | (seems to be the max), shouldn't that give you plenty of
         | ability to run the largest models (albeit a bit slowly).
         | 
         | It seems that AMD Epyc CPUs support terabytes of ram, some are
         | as cheap as 1000 EUR. why not just run the full R1 model on
         | that - seems that it would be much cheaper than multiple of
         | those insane NVidia-Karten.
        
           | throw-qqqqq wrote:
           | The bottleneck is mainly memory bandwidth. AMD EPYC hw is
           | appealing for local inference because it has a higher memory
           | bandwidth than desktop gear (because 8-12 memory channels vs
           | 2 on almost everything else), but not as fast as the Apple
           | architectures and nowhere near VRAM speeds. If you want to
           | drastically exceed ~3-5 tokens/s on 70b-q4 models, you
           | usually still need GPUs.
        
             | timschmidt wrote:
             | On Zen5 you also get AVX512 which llamafile takes advantage
             | of for drastically improved speeds during prompt
             | processing, at least. And the 12 channel Epycs actually
             | seem to have more memory bandwidth available than the Apple
             | M series. Especially considering it's all available to the
             | CPU as opposed to just some portion of it.
        
               | Gracana wrote:
               | Maybe EPYC can make better use of the available
               | bandwidth, but for comparison I have a water cooled Xeon
               | W5-3435X running at 4.7GHz all-core with 8 channels of
               | DDR5-6400, and CPU inference is still dog slow. With a
               | 70B Q8 model I get 1 tok/s, which is a lot less than I
               | thought I would get with 410GB/s max RAM bandwidth. If I
               | run on 5x A4000s I get 6.1 tok/s, which makes sense...
               | 448GB/s / 70GB = 6.4 tok/s max.
        
               | iamnotagenius wrote:
               | very strange as I get on old i5-12400+DDR4 2 tok/sec with
               | 14B/q8 model.
        
             | magicalhippo wrote:
             | This was beautifully illustrated in the recent Phoronix
             | 5090 LLM benchmark[1], which I noted here[2]. The tested
             | GPUs had an almost perfect linear relationship between
             | generated token/s and GB/s memory bandwidth, except the
             | 5090 where it dipped slightly.
             | 
             | I guess the 5090 either started ever so slightly to become
             | compute limited as well, or hit some overhead limitation.
             | 
             | [1]: https://www.phoronix.com/review/nvidia-rtx5090-llama-
             | cpp
             | 
             | [2]: https://news.ycombinator.com/item?id=42847284
        
             | jmb99 wrote:
             | It's more expensive, but Zen4 Threadripper Pro is probably
             | the way to go on that front. 8 memory channels, with DIMMs
             | available up to DDR5-7200 for 8x32GB (256GB), or DDR5-6800
             | for 8x48GB (384GB). It'll set you back ~$3k for the RAM and
             | ~$6k for a CPU with 8 CCDs (the 7985WX, at least), and then
             | ~$1k for motherboard and however much you want to spend on
             | NVME. Basically ~$10k for a 384GB DDR5 system with ~435GB/s
             | actual bandwidth. Not quite as fast as the 192GB Apple
             | machines, but twice as much memory and more compute for
             | "only" a few thousand more.
        
               | sourcecodeplz wrote:
               | At these prices, I would just get 2xDigits for $6k and
               | have 256gb.
        
               | jadbox wrote:
               | I have a feeling that Digits will probably get sold out
               | and will pricing will get hiked WAY up.
        
               | riku_iki wrote:
               | is it confirmed that you can get 256gb of vram for that
               | amount? Because my understanding is that digits pricing
               | will start at $3k for some basic config.
        
               | fluoridation wrote:
               | What they meant is buying two whole separate computers.
        
               | riku_iki wrote:
               | I understand. It is still unclear if you can get 128GB
               | vram for $3k.
        
               | fluoridation wrote:
               | Well, I mean, the press release is pretty unambiguous.
               | 
               | >Each Project DIGITS features 128GB of unified, coherent
               | memory and up to 4TB of NVMe storage.
               | 
               | Even if $3k is only the starting price, it doesn't sound
               | like spending more buys you more memory.
        
               | riku_iki wrote:
               | Ok, but it is not clear what kind of RAM is that, how
               | many memory channels, etc. If the goal is to have just
               | 128GB of some ram, then it could be achieved by paying
               | few $100.
        
               | fluoridation wrote:
               | Fine, but at that point you're arguing about the concept
               | of the product. It's billed as a computer for AI and
               | you're saying that it might not be more suitable for AI
               | than a regular PC.
        
               | riku_iki wrote:
               | it is possible that one could build better PC than digits
               | for AI. We will see once they release digits.
        
           | immibis wrote:
           | FWIW Threadrippers go up to 1TB and Threadripper Pro up to
           | 2TB. That's even in the lowest model of each series. (I know
           | this because it happens to be the chip I have. Not saying you
           | shouldn't go for Epyc if it works out better.)
        
         | immibis wrote:
         | The real insult here is graphics card vendors refusing to make
         | ones with more than 24GB for several years now. They do this so
         | you'll have to buy several cards for your AI workstation.
         | Hopefully Apple eating their lunch fixes this.
        
           | TeMPOraL wrote:
           | > _They do this so you 'll have to buy several cards for your
           | AI workstation._
           | 
           | AFAIK you can't do that with newer consumer cards, which is
           | why this became an annoyance. Even a RTX 4070 Ti with its 12
           | GB would be fine, _if you could easily stack a bunch of them_
           | like you used to be able with older cards.
        
             | Gracana wrote:
             | It's "easy" if you have a place to build an open frame rig
             | with riser cables and whatnot. I can't do that, so I'm
             | going the single slot waterblock route, which unfortunately
             | rules out 3090s due to the memory on the back side of the
             | PCB. It's very frustrating.
        
               | diggan wrote:
               | I think parents point is that NVLink no longer ships with
               | consumer cards. Before you could buy two cards + a cable
               | between them, and software can treat them as one card.
               | Today you need software support for splitting between the
               | cards, unless you go for "professional" cards or whatever
               | they call them.
        
               | Gracana wrote:
               | Maybe that's what they meant, and it'd be cool if nvidia
               | still offered that on consumer cards, but thankfully you
               | don't need it for LLM inference. The traffic between
               | cards is very small.
        
               | diggan wrote:
               | Isn't the issue that the software needs to explicitly add
               | support for it now, compared to yester-yesterday when you
               | could just treat them as one in software?
        
               | numpad0 wrote:
               | There was a rumor that 5090 or 5090D for China may or may
               | not come with multi-GPU software locked. I think GP's
               | referring to that. It's not clear if it is the case with
               | retail cards.
        
           | therealpygon wrote:
           | I honestly don't know why people aren't more upset by this
           | and still get on their knees for Nvidia. They made the
           | decision specifically to cripple consumer card memory because
           | they didn't like data centers were using them instead of
           | buying their overpriced enterprise cards that were less
           | performant. They removed NVLink because people were getting
           | better performance out of their two $400 cards than the
           | $1,500 cards Nvidia was trying to peddle. They willfully
           | screw consumers and people love them for it.
        
             | dagaci wrote:
             | Because sensible people just use the cloud at this point,
             | you can probably get several years of training for $6000
        
               | immibis wrote:
               | It buys you approximately two days (with reservation
               | discount) of a single p5.48xlarge instance, which has 2TB
               | of RAM, and 640GB of VRAM in 8x H100 cards. In fact that
               | is the pricing example they use:
               | https://aws.amazon.com/ec2/capacityblocks/pricing/
        
               | dagaci wrote:
               | MI300X (RunPod) 192gb ram Hourly Rate: $2.49/hr. Break-
               | even Point: You can rent for 2,410 hours (~100 days of
               | non-stop-continuous use) before reaching the cost of the
               | $6000 Mac. Mac's top out at 192GB not 2TB ;)
               | Consideration: If your AI training requires sporadic use
               | (e.g., a few hours daily or weekly), renting is
               | significantly cheaper. MI300X will also get you result
               | many times faster too, so you could probably multiply
               | that 100 days!
        
           | regularfry wrote:
           | The 5090 is 32GB out of the box. Not that that's anywhere
           | near the top of what you can do on an Apple, but at least
           | it's movement.
        
           | sliken wrote:
           | Or buy 2 Nvidia digits for $6,000 to get 256GB vram.
        
         | yobid20 wrote:
         | I canceled my OpenAI subscription last night, as did many many
         | others. There were some threads in reddit with everyone chiming
         | in they all just canceled too. imo OpenAI is done, and will go
         | through massive cuts and probably acquired by the end of the
         | year for a very tiny fraction of its current value.
        
           | Voloskaya wrote:
           | You want to bet? The panic around deepseek is getting
           | completely disconnected from reality.
           | 
           | Don't get me wrong what DS did is great, but anyone thinking
           | this reshape the fundamental trend of scaling laws and make
           | compute irrelevant is dead wrong. I'm sure OpenAI doesn't
           | really enjoy the PR right now, but guess what
           | OpenAI/Google/Meta/Anthropic can do if you give them a recipe
           | for 11x more efficient training ? They can scale it to their
           | 100k GPUs clusters and still blow everything. This will be
           | textbook Jevons paradox.
           | 
           | Compute is still king and OpenAI has worked on their training
           | platform longer than anyone.
           | 
           | Of course as soon as the next best model is released, we can
           | train on its output and catch up at a fraction of the cost,
           | and thus the infinite bunny hopping will continue.
           | 
           | But OpenAI is very much alive.
        
             | lolinder wrote:
             | > The panic around deepseek is getting completely
             | disconnected from reality.
             | 
             | This entire hype cycle has long been completely
             | disconnected from reality. I've watched a lot of hype
             | waves, and I've never seen one that oscillates so wildly.
             | 
             | I think you're right that OpenAI isn't as hurt by DeepSeek
             | as the mass panic would lead one to believe, but it's
             | _also_ true that DeepSeek exposes how blown out of
             | proportion the initial hype waves were and how inflated the
             | valuations are for this tech.
             | 
             | Meta has been demonstrating for a while that models are a
             | commodity, not a product you can build a business on.
             | DeepSeek proves that conclusively. OpenAI isn't finished,
             | but they need to continue down the path they've already
             | started and give up the idea that "getting to AGI" is a
             | business model that doesn't require them to think about
             | product.
        
               | regularfry wrote:
               | In a sense it doesn't, in that if DeepSeek can do _this_
               | , making OpenAI-type capabilities available for Llama-
               | type infrastructure costs, then if you apply OpenAI scale
               | infrastructure again to a much more efficient
               | training/evaluation system, everything multiplies back
               | up. I think that's where they'll have to head: using
               | their infrastructure moat (such as it is) to apply these
               | efficiency learnings to allow much more capable models at
               | the top end. Yes, they can't sleep-walk into it, but I
               | don't think that was ever the game.
        
             | coliveira wrote:
             | Computing is not king, DeepSeek just demonstrated
             | otherwise. And yes, OpenAI will have to reinvent itself to
             | copy DS, but this means they'll have to throw away a lot of
             | their investment in existing tech. They might recover but
             | it is not a minor hiccup as you suggest.
        
               | nlh wrote:
               | I just don't see how this is true. OpenAI has a massive
               | cash & hardware pile -- they'll adapt and learn from what
               | DeepSeek has done and be in a position to build and train
               | 10x-50x-100x (or however) faster and better. They are
               | getting a wake-up call for sure but I don't think much is
               | going to be thrown away.
        
             | throwup238 wrote:
             | _> The panic around deepseek is getting completely
             | disconnected from reality._
             | 
             | Couldn't agree more! Nobody here read the manual. The last
             | paragraph of DeepSeek's R1 paper:
             | 
             |  _> Software Engineering Tasks: Due to the long evaluation
             | times, which impact the efficiency of the RL process,
             | large-scale RL has not been applied extensively in software
             | engineering tasks. As a result, DeepSeek-R1 has not
             | demonstrated a huge improvement over DeepSeek-V3 on
             | software engineering benchmarks. Future versions will
             | address this by implementing rejection sampling on software
             | engineering data or incorporating asynchronous evaluations
             | during the RL process to improve efficiency._
             | 
             | Just based on my evaluations so far, R1 is not even an
             | improvement on V3 in terms of real world coding problems
             | because it gets stuck in stupid reasoning loops like
             | whether "write C++ code to ..." means it can use a C
             | library or has to find a C++ wrapper which doesn't exist.
        
             | miroljub wrote:
             | > You want to bet?
             | 
             | Why would anyone bet? They can just short the OpenAI / MS
             | stocks, and see in a few months if they were right or not.
        
               | icedchai wrote:
               | 1) OpenAI isn't public, so not possible. 2) MS is one of
               | the most well diversified tech companies, so, if
               | anything, this will be a positive.
        
               | IAmGraydon wrote:
               | OpenAI isn't publicly traded and MSFT's stake is so minor
               | compared to their other business that it will have a
               | negligible impact on their stock price.
        
               | skeaker wrote:
               | How is that any different from a bet?
        
             | dimgl wrote:
             | Deepseek is not the only reason. I cancelled my OpenAI
             | subscription because I've replaced it wholesale with
             | Anthropic.
        
           | kebaman wrote:
           | Doesn't Microsoft own 49% of OpenAI? They'll end up with it
           | all as a division of Microsoft.
        
             | wqaatwt wrote:
             | I think they "own" 49% of OpenAI's net income until a
             | certain very high amount. Not a share of the actual
             | company.
        
               | nicce wrote:
               | They "own" even 75% of profits until Microsoft has
               | recouped its $13 billion investment. 49% comes after
               | that.
        
           | generalizations wrote:
           | In my experience with deepseek and o1, openai's big talk
           | about (and investment into) hallucination avoidance might
           | save their hides here. Deepseek may be smarter, and
           | understand complex problems better, but it also seems to make
           | mistakes more often. (It's as if it's comprehension is
           | better, but it's worse at memorization/recall.)
           | 
           | Need an LLM to one-shot some complex network scripting? as of
           | last night, o1 is still where its at.
        
             | flir wrote:
             | My experience gels with yours. Given the same code sample,
             | DeepSeek has better, more creative suggestions about how to
             | improve it, but it can't implement them without breaking
             | the code. o1, generally, can implement DeepSeek's
             | suggestions successfully. I think chaining them together
             | might have quite interesting results.
        
               | wordpad25 wrote:
               | Is there a tool that can automate chaining like that?
        
               | throwup238 wrote:
               | Aider has an architect mode where it asks one model to
               | plan out the changes and another to actually write the
               | code.
        
               | manmal wrote:
               | I've used it today, with R1 as architect and Sonnet as
               | editor model. So far, this works great. There's no need
               | to use a reasoning model as editor IMO.
               | 
               | Alex (https://alexcodes.app) also does this now btw.
        
             | HarHarVeryFunny wrote:
             | That's ok if all you want to know is which model should I
             | use today, but a test like that is totally dependent on
             | training data, and there is no reason to expect that either
             | DeepSeek-V3 (the base model for R1) or the additional
             | training data for R1 is that same as what OpenAI used for
             | O1 and whatever base model it was built on.
             | 
             | The benchmark comparisons are perhaps, for now, the best
             | way to compare reasoning prowess of R1 vs O1, since it
             | seems pretty certain they both trained for those cases.
             | 
             | I think the real significance of R1 isn't the released
             | model/weights itself, but more the paper detailing (sans
             | training data) how to replicate it, and how effective
             | "distillation" (i.e. generate synthetic reasoning data for
             | SFT) can be to enhance reasoning even without using RL.
        
           | wqaatwt wrote:
           | IMHO o1 it's still comparable to a lot better for
           | accomplishing actual stuff than DeepSeek. At least for my use
           | cases.
           | 
           | Of course cost is incomparably higher since plus has a very
           | low limit. Which of course is a huge deal.
        
           | conradfr wrote:
           | Why every time there is a new model all the other competitors
           | are declared immediately dead?
        
             | osigurdson wrote:
             | Because we like drama.
        
             | HarHarVeryFunny wrote:
             | The big deal here isn't that R1 makes any other models
             | obsolete in terms of performance, but how cheap it is $2 vs
             | $60 per million output tokens compared to O1 (which it
             | matches in benchmark performance).
             | 
             | O1 vs R1 performance on specific non-benchmark problems is
             | also not that relevant until people have replicated R1
             | and/or tried fine-tuning it with additional data. What
             | would be interesting to see is whether (given the different
             | usage of RL) there is any difference in how well R1 vs O1
             | generalize to reasoning capability over domains they were
             | not specifically trained for. I'd expect that neither do
             | that well, but not knowing details of what they were
             | trained on makes it hard to test.
        
           | yieldcrv wrote:
           | I disagree, I don't really need "conversational chat
           | responses", I need multimodal
           | 
           | ChatGPT is the king of the multimodal experience still.
           | Anthropic is a distant second, only because it lets you
           | upload images from the clipboard and responds to them, but it
           | can't do anything else like generate images - sometimes it
           | will do a flowchat which is kind of cool, GPT won't do that -
           | but will it speak to you, have tones, listen to you? no.
           | 
           | And in the open source side, this area has been stagnant for
           | like 18 months. There is no cohesive multimodal experience
           | yet. Just a couple vision models with chat capabilities and
           | pretty pathetic GUIs to support them. You have to still do
           | everything yourself there.
           | 
           | There is a huge utility for me, and many others that dont
           | know it yet, if we could just load a couple models at once
           | that work together seamlessly in a single seamless GUI like
           | how ChatGPT works.
        
           | nicce wrote:
           | 1. You can get all the models by buying Kagi subscription
           | (excluding o1). Includes DeepSeek models. You can also feed
           | the assistant with search data that you can filter.
           | 
           | 2. If you have GitHub Copilot, you get o1 chat also there.
           | 
           | I haven't seen much value with OpenAI subscription for ages.
        
         | mory91 wrote:
         | idk, in my daily work i still see o1 being more useful, did you
         | observe both having the same reasoning power?
        
       | miohtama wrote:
       | Flappy Bird in Python is the new Turing test
        
         | danielhanchen wrote:
         | :) It's my goto test :) I did amp it up by adding 10 conditions
         | and made a scoring card - I found the original R1 to sometimes
         | forget "import os" or miss some lines as well, so I thought it
         | was at least a good check!
         | 
         | I also like to ask the models to create a simple basic
         | Minecraft type game where you can break pieces and store them
         | in your inventory, but disallow building stuff
        
           | miohtama wrote:
           | I feel any AI can fix those problems when they can finally
           | act. The problem AIs cannot run or debug code, or even book a
           | hotel for me. When that is solved and an AI can interact with
           | the code like a human does, it can fix its problems like a
           | human does.
        
             | merman wrote:
             | Exactly! Why can't LLMs run their own code?
        
               | Applejinx wrote:
               | Rampancy.
        
               | whimsicalism wrote:
               | they can, feel free to inference and give it an
               | interpreter
        
         | mclau156 wrote:
         | hopefully we eventually push them to make more classic games
         | like motherlode
        
       | brap wrote:
       | As someone who is out of the loop, what's the verdict on R1? Was
       | anyone able to reproduce the results yet? Is the claim that it
       | only took $5M to train generally accepted?
       | 
       | It's a very bold claim which is really shaking up the markets, so
       | I can't help but wonder if it was even verified at this point.
        
         | huijzer wrote:
         | > Is the claim that it only took $5M to train generally
         | accepted?
         | 
         | Based on Nvidia being down 18% yesterday I would say the claim
         | is generally accepted.
        
           | deskamess wrote:
           | > Nvidia being down 18%
           | 
           | The only part of DeepSeek-R1 I do not like. I hope it's over,
           | but I am not holding my breath.
        
             | coffeebeqn wrote:
             | Nvidia is now up only 1906% over 5 years. What a disgrace
        
               | samvher wrote:
               | It crashed all the way back to June 2024 levels, eons of
               | progress wiped out
        
           | tarruda wrote:
           | It is still unconfirmed since no one outside of deepseek
           | reproduced it.
           | 
           | If confirmed, Nvidia could go down even more
        
             | Wheaties466 wrote:
             | based on information and background they thoroughly gave
             | when releasing their research its pretty easy to put
             | together that it did take them significantly less resources
             | to train this model. only having specific parameters
             | available at a time instead of activating everything all at
             | once is pretty ingenious.
             | 
             | that and they just happened to be undergoing a large scale
             | "cyber attack"
        
             | Majromax wrote:
             | I'm not sure I see the bear argument for NVidia here. Huge
             | AI models certainly drive NVidia sales, but huge AI models
             | are also widely thought to be untrainable and nearly un-
             | runnable save for large datacenters.
             | 
             | To me, this is ripe for an application of the Jevons
             | paradox. If architectural improvements make similar models
             | cheaper, I would expect to see _more_ of them trained and
             | deployed, not fewer, ultimately increasing the market for
             | GPU-like hardware.
        
           | willsmith72 wrote:
           | Because the markets are rational, all-knowing, and have never
           | been wrong?
        
             | huijzer wrote:
             | That was not the question.
        
               | mcv wrote:
               | It is if you're using market movements as evidence of
               | anything factual. If markets aren't rational, you can't
               | use them that way.
        
               | whimsicalism wrote:
               | do you only take advice/learn from all-knowing people?
        
               | mcv wrote:
               | Do you know any?
               | 
               | But here's my advice: drop the fallacious arguments and
               | try something more honest.
        
               | whimsicalism wrote:
               | my argument isn't fallacious - it is logical: we can
               | learn/use evidence from something without presuming it is
               | all knowing. you are putting words in others mouths that
               | they did not say
        
               | mcv wrote:
               | I'm sorry, I thought you introduced the "all-knowing" out
               | of nowhere, but this was indeed mentioned by willsmith72.
               | I'd missed that.
               | 
               | Still, his implied assertion that markets that markets
               | can often behave irrationally, and can't be used as
               | evidence of technical matters, seems pretty valid to me.
               | 
               | But I suppose you could see it as a sign that something
               | is at least temporarily "generally accepted" among
               | investors. That doesn't mean it's generally accepted
               | among AI researchers, though.
               | 
               | Although I thought it was $6M rather than $5M, and that
               | that was only the last step, and not the total
               | investment. What does seem to be generally accepted among
               | investors that this isn't good news for NVidia's profits,
               | but that still doesn't mean that all the specific facts
               | are generally accepted.
        
               | tucnak wrote:
               | Don't worry, NVDA will bounce back and you will get a
               | chance to get out.
               | 
               | Efficient market hypothesis is for nerds!
        
             | pgwhalen wrote:
             | No, because the market is an aggregate of opinions, so it's
             | entirely fair to say it's "generally accepted." That has
             | nothing to do with whether something happens to be true or
             | not.
             | 
             | It may provide a financial opportunity for someone who
             | disagrees with that aggregated opinion though.
        
             | ahmeneeroe-v2 wrote:
             | as opposed to HN comments??
        
           | infecto wrote:
           | While Deepseek was an instigator in the price movements I
           | would not say its accepted.
        
           | afavour wrote:
           | I don't see them as related. The market moves when there is
           | money to be made. It's only tangentially related to any kind
           | of general sentiment.
           | 
           | "I don't believe this, but I know others will, so I'm
           | selling"
        
         | Kye wrote:
         | Huggingface is working on reproducing it:
         | https://github.com/huggingface/open-r1
        
         | infecto wrote:
         | I think the jury is out. With folks trying to replicate the
         | process we will see if the low budget is true or not. I am
         | still on the fence, there was comments from Scale CEO that they
         | have a huge number of H100s they used. On the market side I
         | think regardless if this was true or not, this gave people the
         | opportunity to sell what is perhaps overinflated valuations.
        
         | dinosaurdynasty wrote:
         | That's likely only the marginal cost of training this model,
         | and doesn't include a lot of other costs, like the datacenters
         | and GPUs themselves which they already had and also the staff.
         | 
         | If they aren't lying because they have hardware they're not
         | supposed to have, which is also a possibility.
        
           | whimsicalism wrote:
           | these claims are getting more wrong every time i see them,
           | weird game of telephone going around tech circles.
           | 
           | the cost absolutely includes the cost of GPUs and data
           | centers, they quoted a standard price for renting h800 which
           | has all of this built in. but yes, as very explicitly noted
           | in the paper, it does not include cost of test iterations
        
         | whimsicalism wrote:
         | r1 probably cost way less to train, $5m is the alleged price
         | tag for dsv3
        
         | ryao wrote:
         | They claimed that it only took $5 million of GPUs to train
         | Deepseek v3, which was the base model. They did not claim that
         | the total costs were $5 million. They omitted the costs of
         | additional hardware, electricity, personnel, training dataset
         | acquisition, etcetera. They likely spent tens of times more on
         | this at a minimum.
         | 
         | That said, what they did with $5 million of GPUs is impressive.
         | Reportedly, they resorted to using PTX assembly to make it
         | possible:
         | 
         | https://www.tomshardware.com/tech-industry/artificial-intell...
        
       | mtrovo wrote:
       | Wow, an 80% reduction in size for DeepSeek-R1 is just amazing!
       | It's fantastic to see such large models becoming more accessible
       | to those of us who don't have access to top-tier hardware. This
       | kind of optimization opens up so many possibilities for
       | experimenting at home.
       | 
       | I'm impressed by the 140 tokens per second speed with the
       | 1.58-bit quantization running on dual H100s. That kind of
       | performance makes the model practical for small or mid sized
       | shops to use it for local applications. This is a huge win for
       | people working on agents that require low latency that only local
       | models could support.
        
         | danielhanchen wrote:
         | I was pleasantly surprised by 140 tokens/s as well! I literally
         | thought I did something wrong but it was real!
        
         | paradite wrote:
         | Btw completely off topic, but your comment triggered the
         | internal classification in my brain, and it looks like AI-
         | generated.
         | 
         | Not accusing you anything. Could be that you happen to write in
         | a way similar to LLMs. Could be that we are influenced by LLM
         | writing styles and are writing more and more like LLMs. Could
         | be that the difference between LLM generated content and human-
         | generated content is getting smaller and harder to tell.
        
           | j_bum wrote:
           | +1 my LLM spidy senses were tingling.
           | 
           | It's the exclamation point in the first paragraph, the
           | concise and consistent sentence structure, and the lack of
           | colloquial tone.
           | 
           | OP, no worries if you're real. I often read my own messages
           | or writing and worry that people will think I'm an LLM too.
        
             | jasonjmcghee wrote:
             | "This kind of optimization opens up so many possibilities"
             | was what triggered me.
        
           | mtrovo wrote:
           | haha you got me. I'm real person using LLM to proofread the
           | stuff I write. English is not my native language and I'm
           | trying to improve my written vocabulary a little bit. Sorry
           | if it reads a little bit too off.
        
             | paradite wrote:
             | Haha no worries. This is a perfectly valid use case of LLM.
             | I'm happy that the comment sounds very professional and to
             | the point.
        
           | ahmeneeroe-v2 wrote:
           | Very funny, I didn't mentally jump to LLM, but the language
           | was so lifeless that I stopped reading.
           | 
           | Amazing that OP confirmed you're correct (and good use of LLM
           | @OP).
        
       | CodeCompost wrote:
       | Can I run this on ollama?
        
         | benoitg wrote:
         | Yes, the instructions are in the OP.
        
       | Jasondells wrote:
       | An 80% size reduction is no joke, and the fact that the 1.58-bit
       | version runs on dual H100s at 140 tokens/s is kind of mind-
       | blowing. That said, I'm still skeptical about how practical this
       | really is for most people. Like, yeah, you can run it on 24GB
       | VRAM or even with just 20GB RAM, but "slow" is an understatement
       | --those speeds would make even the most patient person throw
       | their hands up.
       | 
       | And then there's the whole repetition issue. Infinite loops with
       | "Pygame's Pygame's Pygame's" kind of defeats the point of
       | quantization if you ask me. Sure, the authors have fixes like
       | adjusting the KV cache or using min_p, but doesn't that just
       | patch a symptom rather than solve the actual problem? A fried
       | model is still fried, even if it stops repeating itself.
       | 
       | On the flip side, I love that they're making this accessible on
       | Hugging Face... and the dynamic quantization approach is pretty
       | brilliant. Using 1.58-bit for MoEs and leaving sensitive layers
       | like down_proj at higher precision--super clever. Feels like
       | they're squeezing every last drop of juice out of the
       | architecture, which is awesome for smaller teams who can't afford
       | OpenAI-scale hardware.
       | 
       | "accessible" still comes with an asterisk. Like, I get that
       | shared memory architectures like a 192GB Mac Ultra are a big
       | deal, but who's dropping $6,000+ on that setup? For that price,
       | I'd rather build a rig with used 3090s and get way more bang for
       | my buck (though, yeah, it'd be a power hog). Cool tech--no doubt
       | --but the practicality is still up for debate. Guess we'll see if
       | the next-gen models can address some of these trade-offs.
        
         | danielhanchen wrote:
         | Oh the repetition issue is only on the non dynamic quants :) If
         | you do dynamic quantization and use the 1.58bit dynamic
         | quantized model the repetition issue fully disappears!
         | 
         | Min_p = 0.05 was a way I found to counteract the 1.58bit model
         | generating singular incorrect tokens which happen around 1
         | token per 8000!
        
           | smcleod wrote:
           | min_p is great, do you apply a small amount of temperate as
           | well?
        
             | danielhanchen wrote:
             | The recommended temperature from DeepSeek is 0.6 so I leave
             | it at that!
        
               | smcleod wrote:
               | I think most of the model creators share their model
               | usage examples so high at 0.6-0.7 simply because it's
               | what a lot of the client apps use. IMO this is WAY too
               | high unless you're doing creative writing.
               | 
               | Generally I set temp to 0-0.4 at absolute most.
               | 
               | min_p actually needs a little temperature to work
               | effectively so with min_p I almost always use 0.2
        
               | danielhanchen wrote:
               | Ye lower temp is also good :) Tbh its all trial and error
               | - I found temp=1.5, min_p=0.1 to be very useful for
               | pass@k type workloads - ie calling the LLM multiple times
               | and aggregating.
               | 
               | temp=0 is also good for singular outputs. For
               | classification tasks, it's better to actually inspect the
               | logits.
               | 
               | But my goto setting is always setting min_p at least 0.01
               | or 0.05! It vastly suppresses incorrect rare random
               | tokens from being created, and it helps massively!
        
             | Der_Einzige wrote:
             | Btw, min_p (the paper about the sampler) got accepted to
             | ICLR! As 4th author it warms my heart to so it used so much
             | in the wild.
        
               | danielhanchen wrote:
               | Oh hi!! Congratulations on ICLR!!! min_p = 0.1 and temp =
               | 1.5 is my default goto settings!!
        
         | elorant wrote:
         | Not everyone needs the largest model. There are variations or
         | R1 with fewer parameters that can easily run on consumer
         | hardware. With 80% size reduction you could run 70B on 8-bit on
         | an RTX 3090.
         | 
         | Other than that, if you really need the big one you can get six
         | 3090s and you're good to go. It's not cheap, but you're running
         | a ChatGPT equivalent model from your basement. A year ago this
         | was a wetdream for most enthusiasts.
        
           | Kye wrote:
           | I ran whatever version Ollama downloaded on a 3070ti (laptop
           | version). It's reasonably fast. Generative stuff can get
           | weird if you do prompts like "in the style of" or "a new
           | episode of" because it doesn't seem to have much pop culture
           | in its training data. It knows the Stargate movie, for
           | example, and seems to have the IMDB info for the series, but
           | goes absolutely ham trying to summarize the series.
           | 
           | This line in the stuff inside the <think> section suggests
           | it's also been trained on YouTube clips:
           | 
           | >> _" I'm not entirely sure if I got all the details right,
           | but this is what I remember from watching clips and summaries
           | online."_
           | 
           | An excerpt from the generated summary:
           | 
           | >> _" Set in the 23rd century during a Z-Corp invasion, the
           | series features action sequences, strategic thinking, and
           | humor. It explores themes of international espionage, space
           | warfare, and humanity's role in the cosmos. The show
           | incorporates musical numbers and catchy theme songs for an
           | engaging viewing experience. The plot involves investigating
           | alien warships and their secret base on Kessari planet while
           | addressing personal conflicts and philosophical questions
           | about space."_
           | 
           | "It explores themes of international espionage, space
           | warfare, and humanity's role in the cosmos" is the closest to
           | correct line in the whole output.
        
             | Mashimo wrote:
             | > ran whatever version Ollama downloaded on a 3070ti
             | (laptop version). It's reasonably fast.
             | 
             | Probably was not r1, but one of the other models that got
             | trained on r1, which apparently might still be quite good.
        
               | Kye wrote:
               | I'm not too hip to all the LLM terminology, so maybe
               | someone can make sense of this and see if it's r1 or
               | something based on r1:
               | 
               | >>> /show info                 Model
               | architecture        qwen2              parameters
               | 7.6B              context length      131072
               | embedding length    3584              quantization
               | Q4_K_M
        
               | Mashimo wrote:
               | "Qwen2.5 is the large language model series developed by
               | Qwen team, Alibaba Cloud."
               | 
               | And I think they, the DeepSeek team, finetunes Qwen 7b on
               | DeepSeek. That is how I understood it.
               | 
               | Which apparently makes it quite good for a 7b model. But,
               | again: if I understood it correctly, is still just qween
               | and without the reasoning of DeepSeek.
        
               | randomifcpfan wrote:
               | In my application, code generation, the distilled
               | DeepSeek models (7B to 70B) perform poorly. They imitate
               | the reasoning of the r1 model, but their conclusions are
               | not correct.
               | 
               | The real r1 model is great, better than o1, but the
               | distilled models are not even as good as the base models
               | that they were distilled from.
        
               | whimsicalism wrote:
               | it's a distill, it's going to be much much worse than r1
        
               | woadwarrior01 wrote:
               | Ollama has been deliberately misrepresenting R1 distill
               | models as "R1" for marketing purposes. A lot of "AI"
               | influencers on social media are unabashedly doing the
               | same. Ollama's default "R1" model is a 4-bit RTN
               | quantized 7B model, which is nowhere close to the real R1
               | (a 671B parameter fp8 MoE).
               | 
               | https://www.reddit.com/r/LocalLLaMA/comments/1i8ifxd/olla
               | ma_...
        
               | wklauss wrote:
               | Ollama is pretty clear about it, it's not like they are
               | trying to deceive. You can also download the 671B model
               | with Ollama, if you like.
        
               | woadwarrior01 wrote:
               | Yeah, they're so clear in fact that they call the
               | distilled models "R1" in the url and everywhere on the
               | page[1], instead of using the "DeepSeek-R1-Distill-"
               | prefix, as DeepSeek themselves do[2].
               | 
               | [1]: https://ollama.com/library/deepseek-r1
               | 
               | [2]: https://github.com/deepseek-
               | ai/DeepSeek-R1#deepseek-r1-disti...
        
               | whimsicalism wrote:
               | no they are not, they intentionally remove every
               | reference to this not being r1 from the cli and changed
               | the names from the ones both Deepseek and Huggingface
               | used.
        
               | nyrikki wrote:
               | They used short strings for the names, which is very
               | different from deception.
               | 
               | https://ollama.com/search
               | 
               | > DeepSeek's first-generation of reasoning models with
               | comparable performance to OpenAI-o1, including six dense
               | models distilled from DeepSeek-R1 based on Llama and
               | Qwen.
               | 
               | Well I guess if you are in the Enterprise Java naming
               | model you would expect something like "VisitorModelUtilsL
               | istGetterAdapterInterceptorMessageManagerDrivenObserverPo
               | ol"
               | 
               | If you look at their API docs you will see:
               | model: name of the model to push in the form of
               | <namespace>/<model>:<tag>
               | 
               | I don't think there is any reason to jump to the
               | conclusion it is some type of conspiracy here, just
               | naming things based on a API that probably didn't think
               | about distillation when they created it.
        
               | horsawlarway wrote:
               | I mean... yes. The DeepSeek announcement puts R1 right
               | there in the name for those models. https://api-
               | docs.deepseek.com/news/news250120
               | 
               | It's fairly clear that R1-Llama or R1-Qwen is a distill,
               | and they're all coming directly from DeepSeek.
               | 
               | As an aside, at least the larger distilled models (I'm
               | mostly running r1-llama-distill-70b) are definitely not
               | the same thing as the base llama/qwen models. I'm getting
               | better results locally, admittedly with the slower
               | inference time as it does the whole "<think>" section.
               | 
               | Surprisingly - The content in the <think> section is
               | actually quite useful on its own. If you're using the
               | model to spitball or brainstorm, getting to see it do
               | that process is just flat out useful. Sometimes more-so
               | than the actual answer it finally produces.
        
             | narrator wrote:
             | It is hilariously bad at writing erotica when I've used
             | jailbreaks on it. It's knowledge is the equivalent of a
             | 1980s college kid with no access to pornography who watched
             | an R rated movie once.
        
               | svachalek wrote:
               | That's like trying to assemble an Ikea bookshelf with a
               | bulldozer. All that extra power is doing nothing for the
               | task you're asking of it, and there are plenty of
               | lightweight alternatives.
        
           | brookst wrote:
           | Or if you want a large model but don't need high performance,
           | get a Mac with 128GB UMA.
        
             | hdjjhhvvhga wrote:
             | How many tokens/s would you get in such a setup?
        
           | whimsicalism wrote:
           | There's a huge difference both in capabilities and in meaning
           | between "variations of r1" and "r1 distill". ollama is
           | intentionally misleading people on this but the distills are
           | much much worse
        
             | thot_experiment wrote:
             | They're really not? Both subjectively and in benchmarks
             | there is no world in which the delta between the models
             | deserves a "much much".
        
         | smcleod wrote:
         | I just ran it up on 48gb (2x 3090) + overflow into CPU RAM and
         | it runs at around 4tk/s (only a little 8k context size though)
         | which while absolutely not something I'd personally use daily -
         | it is actually usable.
        
           | eurekin wrote:
           | I have similar set-up - can you help out with running it? Was
           | it in ollama?
           | 
           | EDIT: It seems that original authors provided a nice write-
           | up:
           | 
           | https://unsloth.ai/blog/deepseekr1-dynamic#:~:text=%F0%9F%96.
           | ..
        
             | smcleod wrote:
             | Yep that's pretty much what I did, their calculation for
             | the layers was slightly off though, I found I could offload
             | an extra 1-2 layers to the GPUs
        
               | danielhanchen wrote:
               | Oh yes I reduced it by 4 for just in case :) I found
               | sometimes the formula doesn't work, so in the worst case
               | -4 was used - glad at least it ran!
        
           | ryao wrote:
           | Which one did you run?
        
         | huijzer wrote:
         | > That said, I'm still skeptical about how practical this
         | really is for most people.
         | 
         | I'm running Open WebUI for months now for me and some friends
         | as a front-end to one of the API providers (deepinfra in my
         | case, but there are many others, see
         | https://artificialanalysis.ai/).
         | 
         | Having 1.58-bit is very practical for me. I'm looking much
         | forward to the API provider adding this model to their system.
         | They also added a Llama turbo (also quantized) a few months
         | back so I have good hopes.
        
           | danielhanchen wrote:
           | Oh I love Open WebUI as well!! But glad to hear the 1.58bit
           | version could be helpful to you!
        
         | ricardobeat wrote:
         | The repetition issue happens on simple quantization, what they
         | are releasing is an approach that fixes that.
        
           | danielhanchen wrote:
           | Yes exactly! I edited the blog post to make the wording a bit
           | better!
        
         | jairuhme wrote:
         | At my work, we self-host some models and have found that for
         | anything remotely similar to RAG or use cases that are very
         | specific, the quantized models have proven to be more than
         | sufficient. This helps us keep them running on smaller infra
         | and generally lower costs
        
           | michaelt wrote:
           | Personally I've noticed major changes in performance between
           | different quantisations of the same model.
           | 
           | Mistral's large 123B model works well (but slowly) at 4-bit
           | quantisation, but if I knock it down to 2.5-bit quantisation
           | for speed, performance drops to the point where I'm better
           | off with a 70B 4-bit model.
           | 
           | This makes me reluctant to evaluate new models in heavily
           | quantised forms, as you're measuring the quantisation more
           | than the actual model.
        
             | sitkack wrote:
             | How are you doing your evals?
             | 
             | Being able to do semantic diffs of the output of the two
             | models should tell you what you need to do.
        
             | danielhanchen wrote:
             | That's a fair point - the trick with dynamic quants is we
             | selectively choose not to quantize many components - ie
             | attention is left at 4 or 6bit, just the MoE parts are
             | 1.5bit (-1, 0, 1)
             | 
             | There are distilled versions like Qwen 1.5, 3, 14, 32,
             | Llama 8, 70, but those are distilled - if you want to run
             | the original R1, then the quants are currently the only
             | way.
             | 
             | But I agree quants do affect perf - hence the trick for
             | MoEs is to not quantize specific areas!
        
         | JKCalhoun wrote:
         | Layman here -- but I am hopeful for 1.58 bit plus custom
         | silicon to be the Holy Grail. I suppose I am setting high
         | expectations on Apple to integrate said in their next "A" chip.
         | 
         | Wishful thinking.
        
           | danielhanchen wrote:
           | Ye a custom chip would be insane! 1.5 bit with a scaling
           | factor seems to be actually usable for MoEs with shared
           | experts!
        
         | rafaelmn wrote:
         | >Like, I get that shared memory architectures like a 192GB Mac
         | Ultra are a big deal, but who's dropping $6,000+ on that setup?
         | 
         | AMD strix halo APU will have quad channel memory and will
         | launch soon so expect these kinds of setups available for much
         | less. Apple is charging an arm and a leg for memory upgrades,
         | hopefully we get competition soon. From what I saw at CES OEMs
         | are paying attention to this use case as well - hopefully not
         | following suite on RAM markups.
        
           | dagmx wrote:
           | Unfortunately, Apple's RAM and Storage upgrade prices are
           | very in line with other class comparable OEMs.
           | 
           | I'm sure there'll be some amount of undercutting but I don't
           | think it'll be a huge difference on the RAM side itself.
        
             | rafaelmn wrote:
             | Highly depends on how accessible AMD makes these boards, if
             | a lot of OEMs get it there will be good deals for sure.
             | DDR5 prices are nowhere near Apple markups.
        
             | nyrikki wrote:
             | The upgrade from 8GB to 16GB cost $200...in 2012!!!!
             | 
             | The newest Lenovo workstations that use LPDDR5x with
             | 7467MT/s gets you a 16GB jump for the price apple charges
             | for an 8GB.
             | 
             | 2:1 isn't "class comparable" IMHO.
        
           | sliken wrote:
           | Keep in mind the strix halo APU has a 256 bit wide memory bus
           | and the Mac Ultra has a 1024 bit wide memory bus.
           | 
           | Here's hoping the Nvidia Digit (GB10 chip) has a 512 bit or
           | 1024 bit wide interface, otherwise the Strix Halo will be the
           | best you can do if you don't get the Mac Ultra.
        
             | rafaelmn wrote:
             | I mean it remains to be seen if it will be compute or
             | bandwidth bound, I am sure mac ultra will also have double
             | or triple compute as well.
             | 
             | But in either case its going to do much better than
             | currently available CPUs with easily upgradeable ram. I
             | would not be surprised to see 128gb configurations for
             | around 3k (going of the ASUS g13 announced pricing of
             | arround 2k for 32gb version and them saying it will go up
             | to 128gb).
             | 
             | At that point sure it might not compete with max but its at
             | a much more acceptable price point, it will not be a device
             | you get just for the AI, but a mobile workstation that you
             | can also run some local models on for normal money. Will
             | need to wait and see. I know I am not buying anything from
             | ASUS either way.
        
         | F7F7F7 wrote:
         | People would only be 'throwing their hands up' because
         | commercial LLMs have set unreasonable expectations for folks.
         | 
         | Anyone who has a/the need for or understands the value of a
         | local LLM would be OK with this kind of output.
        
           | bnchrch wrote:
           | Everyone has the need for on device LLM, if the response rate
           | was fast!
        
             | ddingus wrote:
             | I have MLCCHAT on my old Note 9 phone. It is actually still
             | a great phone, but has 5GB RAM. Running an on device model
             | is the first and only use case the RAM actually matters.
             | 
             | And it has a headphone jack, OK? I just hate Bluetooth
             | earbuds. And yeah, it isna problem, but I digress.
             | 
             | When I run a 2.5B model, I get respectable output. Takes a
             | minute or two to process the context, then output begins at
             | somewhere on the order of 4 to 10 tokens per sec.
             | 
             | So, I just make a query and give it a few and I have my
             | response.
             | 
             | Here is how I see it:
             | 
             | That little model, which is Gemma 2.2b sorry, knows a lot
             | of stuff. It has knowledge I don't and it gives it to me in
             | a reasonable, though predictable way. Answers are always of
             | a certain teacher reminding student how it all goes way.
             | 
             | I don't care. Better is nice, but if I were stuck somewhere
             | with no network, being able to query that model is amazing!
             | 
             | First aid, how to make fires, materials and uses. Fixing
             | stuff, theories of operation, what things mean and more are
             | in that thing ready for me to take advantage of.
             | 
             | I consider what I have fast. And it will get one or two
             | orders faster over the next few years too.
             | 
             | I did it on a lark (ask the model what that means) and was
             | surprised to see I gained a nice tool.
        
               | jpc0 wrote:
               | > First aid, how to make fires, materials and uses
               | 
               | This scares me more than it should...
               | 
               | Please do not trust an AI in actual life and death
               | situations... Sure if it is literally your only option,
               | but this implies you have a device on you that could make
               | a phone call to an emergency number where a real human
               | with real training and actually correct knowledge can
               | assist you.
               | 
               | Even as an avid hiker the amount of times I've been out
               | off cell service is miniscule and I absolutely refresh my
               | knowledge on first aid regularly and any potential
               | threats before a hike somewhere new.
        
             | ddingus wrote:
             | I fired the model up at 8:39, ready to chat just a bit
             | before 8:40, during which I input the following prompt:
             | 
             | Please define "lark" for me, as in "I did it on a lark."
             | 
             | Output began a bit before 8:41, and was complete before
             | 8:43:
             | 
             | ---Begin Model Output
             | 
             | "On a lark" means without much planning or reason,
             | impulsively or unexpectedly.
             | 
             | Here's how it works:
             | 
             | "I did it on a lark." This means the person acted without a
             | specific plan or reason. "A lark" refers to the lark bird,
             | a small songbird known for its playful nature.
             | 
             | So, "on a lark" suggests something spontaneous, casual, and
             | often with a touch of playful whimsy.
             | 
             | Here are some other ways to understand it:
             | 
             | Without much thought: The person was likely to do it
             | quickly and without much consideration. Based on a whim:
             | It's something they decided to do without a solid plan.
             | 
             | Let me know if you'd like more examples
             | 
             | ---End Model Output
             | 
             | To me, that is pretty damn spiffy!
        
               | dbtc wrote:
               | I opened Dictionary.app and typed in lark and got the
               | definition with etymology in under 30 seconds.
               | 
               | > British English informal used to suggest that an
               | activity is foolish or a waste of time: he's serious
               | about this music lark.
        
           | goosejuice wrote:
           | I use commercial LLMs every day. The best of them can still
           | be infuriating at times to the point of being unproductive.
           | So I'm not sure I agree here.
        
         | sliken wrote:
         | I do want a 192GB Mac Ultra, I'm hoping the Nvidia Digit
         | achieves similar at $3,000. Sadly no specifications or
         | benchmarks, so tokens/sec is just a guess at this point.
        
         | yodsanklai wrote:
         | > I'd rather build a rig with used 3090s and get way more bang
         | for my buck
         | 
         | I'm curious, what would you use that rig for?
        
       | sylware wrote:
       | site is javascript walled
       | 
       | 80%? On 2 H100 only? To get near chatgpt 4? Seriously? The 671B
       | version??
        
         | fsflover wrote:
         | > site is javascript walled
         | 
         | I use Qubes OS to protect myself from the JS.
        
           | sylware wrote:
           | That site should work with a noscript/basic (x)html browser.
        
         | whimsicalism wrote:
         | they have not benchmarked the quantized model.
        
       | CHB0403085482 wrote:
       | DeepSeek R1 in a nutshell
       | 
       | youtube.com/watch?v=Nl7aCUsWykg
        
       | petesergeant wrote:
       | It is going to be truly fucking revolutionary if open-source
       | models are and continue to be able to challenge the state of the
       | art. My big philosophical concern is that AI locks Capital into
       | an absolutely supreme and insurmountable lead over Labour, and
       | into the hands of oligarchs, and the possibility of a future
       | where that's not case feels amazing. It pleases me greatly that
       | this has Trump riled up too, because I think it means he's much
       | less likely to allow existing US model-makers to build moats, as
       | I think he's -- even as a man who I don't think believes in very
       | much -- absolutely unwilling to let the Chinese get the drop on
       | him over this.
        
         | fullstackchris wrote:
         | I have no doubt open source will catch up (it already has, eh?)
         | at the end of the day, it's just creative / new iterations on
         | what is ultimately the transformer architecture... the amount
         | of "secret" moat-like stuff that OpenAI was doing was bound to
         | be figured out or exceeded eventually, like everything in
         | tech...
         | 
         | Not to make fun of OpenAI and the great work they've done but
         | it's kinda like if I went out in the 90s and said I'm going to
         | found a company to have the best REST APIs. You can always
         | found a successful tech company, but you can't found a
         | successful tech company on a technological architecture or
         | pattern alone.
        
       | cubefox wrote:
       | For anyone wondering why "1.58" bits: 2^1.58496... = 3. The
       | weights have one of the three states {-1, 0, 1}.
        
         | dist-epoch wrote:
         | They say something else:
         | 
         | > We managed to selectively quantize certain layers to higher
         | bits (like 4bit), and leave most MoE layers (like those used in
         | GPT-4) to 1.5bit
        
           | cubefox wrote:
           | That was just improper rounding from 1.58 to 1.5. They say
           | 1.58 in other places and explicitly link to
           | https://arxiv.org/abs/2402.17764
        
       | amusingimpala75 wrote:
       | > DeepSeek-R1 has been making waves recently by rivaling OpenAI's
       | O1 reasoning model while being fully open-source.
       | 
       | Do we finally have a model with access to the training
       | architecture and training data set, or are we still calling non-
       | reproducible binary blobs without source form open-source?
        
         | stackedinserter wrote:
         | It sounds like if they owe you the training architecture and
         | training data set.
        
           | chris_pie wrote:
           | It absolutely doesn't. It sounds like further diluting the
           | term "open-source" isn't great.
        
             | cubefox wrote:
             | I assume when people say "open source model" they mean
             | "open weights model". The "open source" term doesn't really
             | make sense here, since machine learning models are not
             | compilations of source code. (Though DeepSeek has published
             | several papers with details on their training process. It's
             | more than just open weights.)
        
               | chris_pie wrote:
               | ML models do have a "source" though
        
               | cubefox wrote:
               | If ML models have a source, brains have a source.
               | 
               | Brains don't have a source.
               | 
               | Therefore, ML models don't have a source.
        
       | afro88 wrote:
       | The size reduction while keeping the model coherent is
       | incredible. But I'm skeptical of how much effectiveness was
       | retained. Flappy bird is well known and the kind of thing a non-
       | reasoning model could het right. A better test would be something
       | off the beaten path that R1 and o1 get right that other models
       | don't.
        
         | whimsicalism wrote:
         | yeah it is pretty unclear how lobotomized it is without
         | benchmark.
         | 
         | i've gotten full fp8 running on 8xh100, probably going to keep
         | doing that
        
       | hendersoon wrote:
       | The size reduction is impressive but unless I missed it, they
       | don't list any standard benchmarks for comparison so we have no
       | way to tell how it compares to the full-size model.
        
       | techwiz137 wrote:
       | How can you have a bit and a half exactly? It doesn't make sense.
        
         | dosinga wrote:
         | It's not a bit and a half. It is 1.58 or really log(3) / log(2)
         | since it allows for three values, -1, 0 an 1
        
       | TheTaytay wrote:
       | Danielhanchen, your work is continually impressive. Unsloth is
       | great, and I'm repeatedly amazed at your ability to get up to
       | speed on a new model within hours of its release, and often fix
       | bugs in the default implementation. At this point, I think
       | serious labs should give you a few hour head start just to iron
       | out their kinks!
        
         | danielhanchen wrote:
         | Oh thanks a lot! Appreciate it :) We're always open to
         | collaborating with anyone!
        
       | bluesounddirect wrote:
       | Hi small comment, please remember in china many things are
       | sponsored by or subsidized by the government. "We[china] can do
       | it for less.." , "it's cheaper in china.." only means the
       | government gave us a pile of cash and help to get here .
       | 
       | I 100% expect some downvotes from the ccp.
        
         | tivert wrote:
         | > Hi small comment, please remember in china many things are
         | sponsored by or subsidized by the government. "We[china] can do
         | it for less.." , "it's cheaper in china.." only means the
         | government gave us a pile of cash and help to get here .
         | 
         | And that's a really important _strategic advantage_ China has
         | versus America, which has such an insane fixation on pure(ish)
         | free markets and free trade that it gives away its advantages
         | in strategic industry after strategic industry.
         | 
         | Some people falsely infer from the experience with the Soviet
         | Union that freer markets always win geopolitical competition,
         | but that's false.
        
           | cynicalpeace wrote:
           | > And that's a really important strategic advantage China has
           | versus America, which has such an insane fixation on
           | pure(ish) free markets and free trade that it gives away its
           | advantages in strategic industry after strategic industry.
           | 
           | > Some people falsely infer from the experience with the
           | Soviet Union that freer markets always win geopolitical
           | competition, but that's false.
           | 
           | The data we have is 500 years of free markets in the western
           | world and the verdict is overwhelmingly: Yes, more freedom
           | means more winning.
           | 
           | Just invite some incompetent bureaucrat over your house to
           | dictate how you should cook and you'll quickly agree.
        
           | syndicatedjelly wrote:
           | It's false except for every time that it has been true
        
         | kccqzy wrote:
         | And the United States subsidizes plenty of things too. For
         | example the CHIPS act has $39 billion in subsidies for chip
         | manufacturing on U.S. soil. There's nothing wrong with either
         | country's subsidies. I personally don't believe in maximum free
         | market. Government subsidy is more often than not a good thing
         | and we need more of them both here and in China.
        
         | lucb1e wrote:
         | > I 100% expect some downvotes from the ccp.
         | 
         | Always happy to oblige when someone insinuates that any critics
         | must be government agents
        
       | MyFirstSass wrote:
       | Is this akin to the quants already being done to various models
       | when you download a GGUF at 4 bits for example, or is this
       | variable layer compression something new that can also be make
       | existing smaller models smaller so we can fit more into say 12 or
       | 16 gb's of vram?
        
       | beernet wrote:
       | Big fan of unsloth, they have huge potential, could definitely
       | need some experienced GTM people though, IMO. The pricing page
       | and messages sent there are really not good.
        
         | danielhanchen wrote:
         | Oh thanks :) Yes agreed we do need better GTM - temporarily
         | it's still me and my brother running Unsloth, so for now we're
         | just prioritizing many more engineering releases :)
        
       | mclau156 wrote:
       | Is the new LLM benchmark to create flappy bird in pygame?
        
       | indigodaddy wrote:
       | Is there any small DS or qwen model that could run on say an M4
       | Mac Mini Standard (16G) ?
        
       | slewis wrote:
       | It would be really useful to see these evaluated across some of
       | the same evals that the original R1 and deepseek's distills were
       | evaluated on.
        
       | danesparza wrote:
       | Just ask it about Taiwan (not kidding). I'm not sure I can trust
       | a model that has such a focused political agenda.
        
       | xiphias2 wrote:
       | Has it been tried on 128GB M4 MacBook Pro? I'm gonna try it, but
       | I guess it will be too slow to be usable.
       | 
       | I love the original DeepSeek model, but the distilled versions
       | are too dumb usually. I'm excited to try my own queries on it.
        
         | prisenco wrote:
         | I'm downloading it now and will report back.
         | 
         | (I've been using the 32B and while it could always be better,
         | I'm not unhappy with it)
        
         | rahimnathwani wrote:
         | I love the original DeepSeek model, but the distilled versions
         | are too dumb usually.
         | 
         | Apart from being dumber, they also don't know as much as R1. I
         | can see how fine-tuning can improve reasoning capability (by
         | showing examples of good CoT) but there's no reason that would
         | improve the knowledge of facts (relative to the Qwen or Llama
         | model on which the finetuning was based).
        
       | Pxtl wrote:
       | Is there any good quick summary of what's special about DeepSeek?
       | I know it's OSS and incredibly efficient, but news laymen are
       | saying it's trained purely on AI info instead of using a corpus
       | of tagged data... which, I assume, means it's somehow extracting
       | weights or metadata or something from other AIs. Is that it?
        
         | rahimnathwani wrote:
         | Is there any good quick summary of what's special about
         | DeepSeek?
         | 
         | Yes, section 2.3 of the Deepseek R1 paper summarizes the
         | training part you're asking about, in less than a page.
         | 
         | https://github.com/deepseek-ai/DeepSeek-R1/blob/main/DeepSee...
        
       | DogRunner wrote:
       | >For optimal performance, we recommend the sum of VRAM + RAM to
       | be at least 80GB+.
       | 
       | Oh nice! So I can try it in my local "low power/low cost" server
       | at home.
       | 
       | My homesystem does run in a ryzen 5500 + 64gb RAM + 7x RTX 3060
       | 12gb
       | 
       | So 64gb RAM plus 84gb VRAM
       | 
       | I dont want to brag around, but point to solutions for us
       | tinkerers with a small budget and high energy costs.
       | 
       | such system can be build for around 1600 euro. The power
       | consumption is around 520 watt.
       | 
       | I started with a AM4 Board (b450 Chipset) and one used RTX 3060
       | 12gb which cost around 200 Euro used if you are patient.
       | 
       | There every additional GPU is connected with the pcie
       | riser/extender to give the cards enough space.
       | 
       | After a while I had replaces the pcie cards with a single pcie x4
       | to 6x PCIe x1 extender.
       | 
       | It runs pretty nice. Awesome to learn and gain experience
        
         | tucnak wrote:
         | How are you arriving at those numbers?
         | 
         | ryzen 5500 + 7x3060 + cooling ~= 1.6 kW off the wall, at 360
         | GB/s memory bandwidth, and considering your lane budget, most
         | of it will be wasted in single PCIe lanes. After-market unit
         | price of 3060's is 200 eur, so 1600 is not good-faith cost
         | estimate.
         | 
         | From the looks of it, your setup is neither low-power, nor low-
         | cost. You'd be better served with a refurbished mac studio
         | (2022) at 400GB/s bandwidth fully utilised over 96 GB memory.
         | Yes, it will cost you 50% more (considering real cost of such
         | system closer to 2000 eur) however it would run at a fraction
         | of power use (10x less, more or less)
         | 
         | I get it that hobbyists like to build PC's, but claiming that
         | sticking seven five year out of date low-bandwidth GPU's in a
         | box is "low power/low cost" is a silly proposition.
         | 
         | You're advocating for e-waste
        
       | ggm wrote:
       | If I invested in a 100x machine because I needed 100 of x to run,
       | and somebody shows how 10x can work, why have I not just become
       | the holder of 10 10x machines, and therefore have already
       | achieved capex to exploit this new market?
       | 
       | I cannot understand why "openai is dead" has legs: repurpose the
       | hardware and data and it can be multiple instances of the more
       | efficient model.
        
       | patleeman wrote:
       | Incredible work by the Unsloth brothers again. It's really cool
       | to see bitnet quantization implemented like this.
        
       | Dwedit wrote:
       | Is this actually 1.58 bits? (Log base 2 of 3) I heard of another
       | "1.58 bit" model that actually used 2 bits instead. "1.6 bit" is
       | easy enough, you can pack five 3-state values into a byte by
       | using values 0-242. Then unpacking is easy, you divide and modulo
       | by 3 up to five times (or use a lookup table).
        
       ___________________________________________________________________
       (page generated 2025-01-28 23:00 UTC)