[HN Gopher] Infrastructure setup and open-source scripts to trai...
       ___________________________________________________________________
        
       Infrastructure setup and open-source scripts to train 70B model
       from bare metal
        
       Author : thejash
       Score  : 291 points
       Date   : 2024-06-27 23:08 UTC (1 days ago)
        
 (HTM) web link (imbue.com)
 (TXT) w3m dump (imbue.com)
        
       | thejash wrote:
       | In the span of a few months, with a small team of researchers and
       | engineers, we trained a 70B parameter model from scratch on our
       | own infrastructure that outperformed zero-shot GPT-4o on
       | reasoning-related tasks. Using our cluster for high performance
       | training meant that every component -- InfiniBand, Ethernet,
       | GPUs, and the nodes themselves -- had to work perfectly. If even
       | a single one of the over 12,000 connections was a little flaky,
       | it could slow down the entire training run.
       | 
       | We're sharing open-source scripts and an end-to-end guide for
       | infrastructure set-up that details the process of making
       | everything work perfectly, and ensuring that it stays that way.
       | 
       | This is one of a three-part toolkit on training a 70b model from
       | scratch. The other two sections focus on evaluations and CARBS,
       | our hyperparameter optimizer; you can find them here:
       | https://imbue.com/research/70b-intro/
       | 
       | Thoughts and questions welcome! :)
        
         | ipsum2 wrote:
         | What happened to the Minecraft-like 3d world your team built?
         | Did you guys pivot?
        
         | Flumio wrote:
         | Nice. Tx for the write up
        
         | chx wrote:
         | > If even a single one of the over 12,000 connections was a
         | little flaky, it could slow down the entire training run
         | 
         | It's an unusual enough sentence to be remarkable and I was like
         | "I read this _exact same_ sentence before ". Indeed, this and
         | most of the writeup appeared on Twitter, LinkedIn, Reddit it
         | seems word-by-word. Is this just spam ?
         | 
         | https://x.com/imbue_ai/status/1805629547473518695
         | 
         | https://reddit.com/r/learnmachinelearning/comments/1dobgbs/t...
         | 
         | https://www.linkedin.com/posts/mattboulos_training-a-70b-mod...
        
           | bottled_poe wrote:
           | lmao, I was thinking this was bullshit and you've cemented
           | that position. We've entered the grifting stage of this AI
           | cycle. Salut.
        
             | knowaveragejoe wrote:
             | Having listened to the person who wrote this speak at
             | length about the subject, it is not BS or grifting.
        
           | leothetechguy wrote:
           | The same company reports multiple times on a finding they've
           | made through multiple social media channels? Shocking. /s
        
           | neilv wrote:
           | I'd rather some company copy&paste the same text multiple
           | places -- if the alternative was that those places would
           | instead get obfuscation of the same information to appear
           | novel each time (so I'd have to read all of them to realize
           | they're all just the same info).
        
           | lolinder wrote:
           | This is the kind of criticism that could only come from
           | someone without much formal writing experience.
           | 
           | This is a very normal workflow: You write a full-length text
           | detailing the project you worked on. You then trim it down to
           | a summary which you share with a group of people X. You then
           | trim it down into a different summary which you share with a
           | group of people Y.
           | 
           | When you do this multiple times you unsurprisingly end up
           | with some sentences that make it into multiple summaries
           | because they're _that_ important to the thesis!
           | 
           | (Also, the summaries on Twitter and Reddit aren't anything
           | close to "most of the writeup"--the full text is 6000+
           | words!)
        
           | ac29 wrote:
           | Eh, seems like legit marketing to me. Yes, they are trying to
           | sell you something, but they are doing that by releasing non-
           | trivial research and open source code.
        
           | fastasucan wrote:
           | I dont inderstand your issue with this. Is it that they share
           | their work several places, or that they don't describe their
           | work in an unique way every time?
        
           | exe34 wrote:
           | i prefer this, to the story about that time they went to
           | Florence and their grandma made pizza for dinner and they got
           | the recipe.
        
         | highfrequency wrote:
         | > outperformed zero-shot GPT-4o
         | 
         | Cool stuff! Does this do RLHF or just pretraining? If the
         | latter, how did you manage to beat GPT 4?
        
         | vessenes wrote:
         | Loved this and the detail - thank you. It's the best inside
         | detail on the engineering work behind these models I've ever
         | read.
         | 
         | Two things I'm curious about- first, what, if any difference
         | would you imagine in training a 400b parameter model? It seems
         | that you have plenty of vram across the cluster, but I want to
         | know what you think.
         | 
         | Second, do you think this sort of architecture is the end game
         | for model training? It seems sooo fragile. Are there better
         | shared training mechanisms/architectures? Are there better
         | cluster geometries?
         | 
         | Thanks again - great read.
        
       | john2x wrote:
       | once the model is trained, what happens to the hardware and
       | infrastructure?
        
         | pvg wrote:
         | It probably isn't the answer but should be - LAN party.
        
           | rvnx wrote:
           | GPUs will be reused for mining Monero and exfiltrate money to
           | the founders at the expense of the investors.
           | 
           | Oops, don't tell I told you.
           | 
           | EDIT: Sorry Dogecoin, thanks to the tip!
        
             | surfingdino wrote:
             | Are you suggesting training models is a cover for mining
             | crypto? The hardware is dual-purpose...
        
             | BetaDeltaAlpha wrote:
             | Monero uses a CPU-optimized consensus algorithm. Dogecoin
             | is a better bet.
        
         | gostsamo wrote:
         | Either training the next model or inference for the already
         | trained one. In some cases, you might even offer it as a
         | service.
        
         | trashtester wrote:
         | Voltage Park is a cloud provider. This is no different from
         | renting barebone infra from AWS, GCP or Azure.
         | 
         | Except Voltage Park, being smaller, is probably more willing to
         | provide some customized setup.
         | 
         | Indeed, they may even see it as a learning opportunity for when
         | they rent similar setups to other customers.
        
       | renewiltord wrote:
       | This is hella cool. Cisco has a new nvidia collab with 800G per-
       | port. I don't recall if it was RoCE or not. The infiniband is
       | accessible by the GPUs here? Beautiful.
       | 
       | Thank you for sharing all this. One of the more directly useful
       | posts.
        
       | lifeisstillgood wrote:
       | I am fascinated by the total electrical power drawn to build
       | models - power and cooling I guess. Do you have any numbers on
       | that (the point being Zuckerberg in a podcast suggested the next
       | 1GW model was being planned - basically a data centre with a mid
       | sized power plant attached)
        
       | instagib wrote:
       | 4,092 H100 GPUs.
       | 
       | They're working on "self-coding". No-code or minimal code
       | solutions or?
       | 
       | Quite a few articles and such people may be interested in also on
       | their website: https://imbue.com/our-work/
        
       | loudmax wrote:
       | This was discussed on the Latent Space podcast a few days ago:
       | https://www.latent.space/p/llm-training-2024
       | 
       | That was a good episode, worth listening to for hearing
       | justifications behind some of these decisions.
        
         | swyx wrote:
         | thank you for listening!
         | 
         | im not used to conducting these kinds of interviews and felt
         | out of my depth. please suggestions questions that you felt
         | should have been asked but werent.
        
       | weinzierl wrote:
       | How much did it cost? Overall, from nothing to the usable model
       | files, in hardware cost, development hours and ultimately
       | electricity and cooling?
        
       | alias_neo wrote:
       | > This post focuses on one cluster that had 4,092 H100 GPUs
       | spread across 511 computers, with eight GPUs to a computer
       | 
       | Am I right in understanding, that's over $100 Million worth of
       | GPUs?
       | 
       | I wonder what/when/if any of this will be within the realms of an
       | enthusiast with a gaming-pc budget.
        
         | freeqaz wrote:
         | Looks correct. They raised $200m from NVIDIA which I presume is
         | in pure GPUs. https://news.crunchbase.com/ai-robotics/new-ai-
         | unicorn-imbue...
        
         | mandeepj wrote:
         | > Am I right in understanding, that's over $100 Million worth
         | of GPUs?
         | 
         | Ha! I guess most or many of the readers (who don't have that
         | much of funding) should jump to the next HN submission
        
       | mmastrac wrote:
       | Honest question: why is there so much PC hardware in the mix
       | here? Why don't we have PCI + infiniband backends with GPUs and a
       | little tiny orchestrating ARM controller and just let them all
       | coordinate with each other? Is it just "momentum" from previous
       | designs and/or lack of "market" for specialized GPU controllers?
        
         | bick_nyers wrote:
         | Are you asking why pay extra for a CPU and RAM? Not everything
         | can be done on a GPU, for example, .png decompression. If you
         | really analyzed your training code and preprocessed your data
         | substantially you could probably get away with very lightweight
         | CPU/RAM resources but I think the reality is that it's such a
         | minor contribution of cost to the overall system (GPU are
         | expensive) that wasting development cycles on that degree of
         | optimization isn't strictly necessary. When you're a
         | hyperscaler you are likely chasing those fractions of a percent
         | of cost efficiency though. To use my original example, you
         | would likely want to preprocess your .png to either .webp
         | (multi-threaded lossless) or .jpeg (lossy), but likely it
         | wouldn't make sense to turn it into a GPU decompressible format
         | as you would save on CPU cost during training but would pay
         | more in storage (and maybe transfer) cost.
         | 
         | Edit: To be more clear, if the CPU work is bottlenecking
         | training, you want to optimize that as much as possible by
         | preprocessing your data/tweaking training scripts. What I'm
         | discussing here is the gap between "fast enough" and "faster":
         | 
         | CPU is not fast enough for training < CPU is exactly fast
         | enough for training < CPU is faster than needed for training
        
         | ianburrell wrote:
         | Cause when you have quarter million dollars of GPU on each
         | machine, it is dumb to worry about few thousand for the
         | controlling hardware. Too risky to use something new.
         | 
         | Another problem is that all the hardware, drivers, and
         | experience for GPU are on PC. It would take a lot of work to
         | get running on ARM since would be starting from scratch. Then
         | more work to get it stable. All to save a little on processor.
        
       | mikewarot wrote:
       | It would be quite interesting to see the same hardware used to
       | repeat the training, but with raw Unicode, instead of tokenized
       | training data.
       | 
       | I'd like to see the difference in performance on spelling and
       | rhymes.
        
       ___________________________________________________________________
       (page generated 2024-06-29 23:02 UTC)