5 ms·
OK, these are a LOT of fun to play with. I've been trying out a quantized version of the Llama 3 one from here: https://huggingface.co/unsloth/DeepSeek-R1-Disti
by simonw 2y ago
OK, these are a LOT of fun to play with. I've been trying out a quantized version of the Llama 3 one from here: https://huggingface.co/unsloth/DeepSeek-R1-Distill-Llama-8B-GGUF https://huggingface.co/unsloth/DeepSeek-R1-Distill-Llama-8B-...
The one I'm running is the 8.54GB file. I'm using Ollama like this:
ollama run hf.co/unsloth/DeepSeek-R1-Distill-Llama-8B-GGUF:Q8_0
You can prompt it directly there, but I'm using my LLM tool and the llm-ollama plugin to run and log prompts against it. Once Ollama has loaded the model (from the above command) you can try those with uvx like this:
uvx --with llm-ollama \
llm -m 'hf.co/unsloth/DeepSeek-R1-Distill-Llama-8B-GGUF:Q8_0' \
'a joke about a pelican and a walrus who run a tea room together'
Here's what I got - the joke itself is rubbish but the "thinking" section is fascinating: https://gist.github.com/simonw/f505ce733a435c8fc8fdf3448e3816b0 https://gist.github.com/simonw/f505ce733a435c8fc8fdf3448e381...
I also set an alias for the model like this:
llm aliases set r1l 'hf.co/unsloth/DeepSeek-R1-Distill-Llama-8B-GGUF:Q8_0'
Now I can run "llm -m r1l" (for R1 Llama) instead.
I wrote up my experiments so far on my blog: https://simonwillison.net/2025/Jan/20/deepseek-r1/ https://simonwillison.net/2025/Jan/20/deepseek-r1/
- reissbaker 2y agoFWIW, you can also try all of the distills out in BF16 on https://glhf.chat https://glhf.chat (either in the UI or via the API), including the 70b. Personally I've been most impressed with the Qwen 32b distill. (Disclosure: I'm the cofounder)
- reissbaker 2y ago(And here's the output of the "pelican and the walrus in a tearoom" joke from DeepSeek's distilled 70b... Mildly better?) <think> Alright, I'm supposed to come up with a joke about a pelican and a walrus in a tea room. Hmm, sounds like a classic setup for a funny situation. Let me think about the characteristics of these animals that could be used humorously. Okay, so pelicans are known for their large beaks, right? They can carry a lot of water in them. Walruses have those big tusks and blubbery bodies. Maybe I can play on the idea of them trying to have a tea party but their physical traits causing some trouble. Wait, being in a tea room, they might be trying to sip tea, but their size or features get in the way. The pelican's beak could be problematic with the cup. And the walrus's size might knock things over or make a mess. Let me consider the flow. Maybe the pelican struggles with the teacup because of his beak, and the walrus has a comment about it. Then the walrus tries to sip but ends up splashing tea everywhere. That could lead to a funny exchange. Putting it together: Pelican says something self-deprecating about his beak not being suited for tea. Then the walrus, who perhaps agrees, tries to sip and splashes tea, making them both laugh. That feels natural and funny. I should make sure the punchline is clear and the dialogue flows smoothly. Maybe start with the pelican struggling, then the walrus makes a comment, and their interaction leads to the humorous conclusion. Yeah, that should work. </think> Why did the pelican and the walrus get kicked out of the tea room? Because the pelican kept dipping his beak in the wrong cup, and the walrus kept breaking the delicate china with his tusks. They just couldn’t “sea” themselves fitting in!
- jpc0 2y agoApparently I'm a sucked for bad jokez... It wasn't that good but the punchline actually got me a giggle...
- oidar 2y agoMight want to let people know that you have to pay to try things ahead of time. It's not obvious when on the launch page.
- reissbaker 2y agoThe launch page has a pretty prominent pricing link — what would make it more obvious?
- oidar 2y agoSomething about pricing above the fold: https://ibb.co/rFR41rK https://ibb.co/rFR41rK
- reissbaker 2y agoGood point! Thank you for the screenshot :)
- widdershins 2y agoYeesh, that shows a pretty comprehensive dearth of humour in the model. It did a decent examination of characteristics that might form the components of a joke, but completely failed to actually construct one. I couldn't see a single idea or wordplay that actually made sense or elicited anything like a chuckle. The model _nearly_ got there with 'krill' and 'kill', but failed to actually make the pun that it had already identified.
- samdjstephens 2y agoYeah it's very interesting... It appears to lead itself astray: the way it looks at several situational characteristics, gives each a "throw-away" example, only to then mushing all those examples together to make a joke seems to be it's downfall in this particular case. Also I can't help but think that if it had written out a few example jokes about animals rather than simply "thinking" about jokes, it might have come up with something better
- linsomniac 2y ago>a joke about a pelican and Tell me you're simonw without telling me you're simonw...
- monkeydust 2y agoThanks! Playing around with this vs the https://ollama.com/tripplyons/r1-distill-qwen-7b https://ollama.com/tripplyons/r1-distill-qwen-7b variant and find 7b to be somewhat of sweet spot of getting to the point with minimal (or less) waffle. Certainly, interesting reading their thought processes, value in that might be greater than the answer itself depending on use-case.
- lmc 2y ago> The walrus might say something like, "We have the biggest catch in town," while the pelican adds a line about not catching any fish recently. It should've stopped there :D
- HarHarVeryFunny 2y agoI think the problem is that humor isn't about reasoning and logic, but almost the reverse - it's about punchlines that surprise us (i.e. not what one would logically anticipate) and perhaps shock us by breaking taboos. Even masters of humor like Seinfeld, with great intuition for what might work, still need to test new material in front of a live audience to see whether it actually does get a laugh or not.
- simonw 2y agoYeah, getting LLMs to invent jokes is an incredibly unfair test of them - but I derive enormous entertainment from seeing them try.
- HarHarVeryFunny 2y agoI just tried Claude Sonnet with the Pelican & Walrus setup, but asked it for something in style of Norm Macdonald's humor, which would seem a potentially good fit for this type of idea. It got the idea of a rambling story (cf Macdonald's moth joke) that mostly ignored them being a Pelican and Walrus, which seemed promising, but still failed to deliver. I'm guessing with some more guidance and iteration it could have come up with something.
- astrange 2y agoI've been trying out "write and perform an MST3K episode reviewing your previous answer" as a prompt. Have yet to get a funny answer from any model. Older base models produce some great stuff though, at least if you're me and think RoboRosewater and dril_gpt2 are funny.
- wat10000 2y agoThis joke is so terrible, I think this might end up being how AI kills us all when it decides it needs us out of the way to make more paperclips.
- stavros 2y agoWait until you see what I got when prompting it for a joke about three priests: > Why did the rabbi become a boxer? > Because he couldn't handle the holy punches! > But wait, that's not quite right. Let me try again. > Here’s one: What do you call a Jewish, Orthodox, and Catholic priest working together? > A trio of men who probably shouldn’t be telling jokes after their day job! > Sorry for the confusion earlier—I hope this one makes more sense! It did not.
- rererereferred 2y ago> What do you call a Jewish, Orthodox, and Catholic priest working together? A miracle.
- wat10000 2y agoThat second one isn’t funny, but it’s still interesting somehow. Actually, it would work as a joke, but it would have to be told by an actual priest hanging out after work with two other actual priests. Ultra-niche humor.
- momojo 2y ago> the joke itself is rubbish but the "thinking" section is fascinating: This is gold. If I was a writer, I'd wring value from that entire thinking-out-loud section and toss the actual punchline. This is weirdly reminiscent of co-programming with CodyAI. It gives me a lot of good 'raw material' and I'm left integrating the last mile stuff.
- _sys49152 2y agolooking at how it thinks, it can help bust you out when you write yourself into a corner.
- BHSPitMonkey 2y agoI like to imagine the entire transcript as a Norm Macdonald bit, a la his moth joke
- simonw 2y agoI got a quantized Llama 70B model working, using most of my 64GB of RAM but it's usable: ollama run hf.co/unsloth/DeepSeek-R1-Distill-Llama-70B-GGUF:Q3_K_M That's a 34GB download. I'm accessing it via https://github.com/open-webui/open-webui https://github.com/open-webui/open-webui which I ran like this: uvx --python 3.11 open-webui serve I have TailScale on my laptop and phone so I can run experiments directly from my phone while leaving my laptop plugged in at home.
- jt_b 2y agoThanks for all your writing on these topics Simon! It has turned me from a bit of a naysayer to an optimist around this tooling, especially being able running stuff locally with access to tools. Have an M4 Pro mac mini arriving this week to have a similar self hosted setup over tailscale.
- brewtide 2y agoMaybe related, but I've also got a tailscale instance running for the same use case (on an older it box but...) I've also installed open-webui attached to ollama. That way the interface I can deal with when on my phone is just a simple nice to use webpage. May want to look into this? Thus far it's worked very slick.
- jt_b 2y agoNice, thanks for the suggestion. I got it set up just before leaving town for a few days, so have been doing a little tinkering with it. I was hoping to have a setup with LM Studio, where my laptop could use the API Server from the mini over the TS network. Unfortunately doesn't seem to be the case, so I'll set up a configuration like you mentioned to just have a global client from any device on the netowrk. It's very cool to be able to have access to such a high horsepower machine from anywhere though. Next step is figuring out the networking interface to be able access the host GPU/ollama API from pods running in a Colima VM/k3s cluster setup.
- matt_daemon 2y ago
- ryanisnan 2y agoSuper interesting. It seems to get hung up on a few core concepts, like the size of the walrus vs. the limited utility of a pelican beak.
- TeMPOraL 2y agoDid you try the universal LLM cheat code as a followup prompt? "Make it better"
- svachalek 2y agoThese "thinking" LLMs are essentially productizing that trick. All the rambling and second guessing they go through at the beginning gives them source material to compute the "better" answer from at the end.
- TeMPOraL 2y agoSure, but you can still re-prompt them again telling them to just do better. In case people missed it, I'm referencing an observation recently made by 'minimaxir, described here: https://minimaxir.com/2025/01/write-better-code/ https://minimaxir.com/2025/01/write-better-code/ As it turns out, you can improve the quality of code generated by some LLMs by repeatedly responding to it with just three words: "write better code". While Max only tested this on Claude 3.5 Sonnet, I see no reason why this wouldn't work with the "thinking" models either. Even if it doesn't the results might still be interesting. With that in mind, here's the article's experiment applied to o1-preview: https://cloud.typingmind.com/share/69e62483-45a4-4378-9915-6029be265c19 https://cloud.typingmind.com/share/69e62483-45a4-4378-9915-6... Eyeballing the output, it seems to align with the article's observation. (o1-preview is the only "thinking" model I currently have API access to; official ChatGPT app doesn't let me reply to o1 family, forcing further interactions to be with "non-thinking" models instead.)
- croemer 2y agoCan someone ELI5 what the difference is between using the "quantized version of the Llama 3" from unsloth instead of the one that's on ollama, i.e. `ollama run deepseek-r1:8b`?
- whereismyacc 2y agoThe weights are quantized down to fewer bits in order to save on memory. The quantization loss is going to result in worse generations.
- ColonelPhantom 2y agoOllama serves multiple versions, you can get Q8_0 from it too: ollama run deepseek-r1:8b-llama-distill-q8_0 The real value from the unsloth ones is that they were uploaded before R1 appeared on Ollama's model list.
- AS04 2y agoUnsloth also works very diligently to find and fix tokenizer issues and many other problems as soon as they can. I have comparatively little trust on ollama following up and updating everything in a timely manner. Last I checked, there is little information on when the GGUFs and etc. on ollama were updated or what llama.cpp version / git commit did they use for it. As such, quality can vary and be significantly lower with the ollama versions for new models I believe.
- dragonwriter 2y agoThey are probably the same model, unsloth does model quants and provides them to the community, AFAIK ollama doesn't, they just indexes publicly available models, whether full or quantized, for convenient use in their frontend.
- tomrod 2y agoCan you recommend hardware needed to run these?
- simonw 2y agoI'm using an M2 64GB MacBook Pro. For the Llama 8B one I would expect 16GB to be enough. I don't have any experience running models on Windows or Linux, where your GPU VRAM becomes the most important factor.
- rane 2y agoWhy isn't GPU VRAM a factor on a Silicon mac?
- Nekit1234007 2y agoBecause there's no VRAM. The “regular” RAM on Apple Silicon devices is shared with the GPU virtually in its entirety.
- dragonwriter 2y agoOn Windows or Linux you can run from RAM or split layers between RAM and VRAM; running fully on GPU is faster than either of those, but the limit on what you can run at all isn’t VRAM.
- akhdanfadh 2y agoSo is it possible to load the ollama deepseek-r1 70b (43gb) model on my 24gb vram + 32gb ram machine? Does this depend on how I load the model, i.e., with ollama instead of other alternatives? Afaik, ollama is basically llama.cpp wrapper. I have tried to deploy one myself with openwebui+ollama but only for small LLM. Not sure about the bigger one, worried if that will crash my machine someway. Are there any docs? I am curious about this and how that works if any.
- yencabulator 2y agoollama runs deepseek-r1:7b on AMD 8945HS, CPU-only, at ~12 tokens/s. You can get started pretty easily in the ~7B model range, for learning purposes.
- fpgaminer 2y agoI think "reasoning" models will solve the joke issue (amongst other issues), but not because they're "reasoning". Rather because they help solve the exploration issue and the scaling issue. Having worked with LLMs a lot for my JoyCaption project, I've got all these hypothesis floating around in my head. I guess the short version, specifically for jokes, is that we lack "joke reasoning" data. The solution, like mathematical problems, is to get the LLM to generate the data and then RL it into more optimal solutions. Longer explanation: Imagine we want an LLM to correctly answer "How many r's are in the word strawberry?". And imagine that language has been tokenized, and thus we can form a "token space". The question is a point in that space, point Q. There is a set of valid points, set A, that encompasses _any_ answer to this question which is correct. There are thus paths through token space from point Q to the points contained by set A. A Generator LLM's job is, given a point, predict valid paths through token space. In fact, we can imagine the Generator starting at point Q and walking its way to (hopefully) some point in set A, along a myriad of inbetween points. Functionally, we have the model predict next token (and hence point in token space) probabilities, and we can use those probabilities to walk the path. An Ideal Generator would output _all_ valid paths from point Q to set A. A Generator LLM is a lossy compression of that ideal model, so in reality the set of paths the Generator LLM will output might encompass some of those valid paths, but it might also encompass invalid paths. One more important thing about these paths. Imagine that there is some critical junction. A specific point where, if the Generator goes "left", it goes into a beautiful flat, grassy plain where the sun is shining. That area is really easy to navigate, and the Generator LLM's predictions are all correct. Yay! But if it goes "right" it ends up in the Fire Swamp with many dangers that it is not equipped to handle. i.e. it isn't "smart" enough in that terrain and will frequently predict invalid paths. Pretraining already taught the Generator LLM to avoid invalid paths to the best of its abilities, but again its abilities are limited. To fix this, we use RL. A Judge LLM takes a completed path and determines if it landed in the set A or not. With an RL algorithm and that reward signal, we can train the Generator LLM to avoid the Fire Swamp, since it often gets low rewards there, and instead goes to the Plain since it often gets rewards there. This results in a Generator LLM that is more _reliable_ and thus more useful. The RL encourages it to walk paths it's good at and capable of, avoid paths it struggles with, and of course encourages valid answers whenever possible. But what if the Generator LLM needs to solve a really hard problem. It gets set down at point Q, and explores the space based on its pretraining. But that pretraining _always_ takes it through a mountain and it never succeeds. During RL the model never really learns a good path, so these tend to manifest as hallucinations or vapid responses that "look" correct. Yet there are very easy, long paths _around_ the mountain that gets to set A. Those don't get reinforced because they never get explored. They never get explored because those paths weren't in the pretraining data, or are so rare that it would take an impractical amount of exploration for the PT model to output them. Reasoning is one of those long, easy paths. Digestible small steps that a limited Generator LLM can handle and use to walk around the mountain. Those "reasoning" paths were always there, and were predicted by the Ideal Generator, but were not explored by our current models. So "reasoning" research is fundamentally about expanding the exploration of the pretrained LLM. The judge gets tweaked slightly to encourage the LLM to explore those kinds of pathways, and/or the LLM gets SFT'd with reasoning data (which is very uncommon in its PT dataset). I think this breakdown and stepping back is important so that we can see what we're really trying to do here: get a limited Generator LLM to find its way around areas it can't climb. It is likely true that there is _always_ some path from a given point Q and set A that a limited Generator LLM can safely traverse, even if that means those paths are very long. It's not easy for researchers to know what paths the LLM can safely travel. So we can't just look at Q and A and build a nice dataset for it. It needs to generate the paths itself. And thus we arrive at Reasoning. Reasoning allows us to take a limited, pretrained LLM, and turn it into a little path finding robot. Early during RL it will find really convoluted paths to the solution, but it _will_ find a solution, and once it does it gets a reward and, hopefully, as training progresses, it learns to find better and shorter paths that it can still navigate safely. But the "reasoning" component is somewhat tangential. It's one approach, probably a very good approach. There are probably other approaches. We just want the best ways to increase exploration efficiently. And we're at the point where existing written data doesn't cover it, so we need to come up with various hacks to get the LLM to do it itself. The same applies to jokes. Comedians don't really write down every single thought in their head as they come up with jokes. If we had that, we could SFT existing LLMs to get to a working solution TODAY, and then RL into something optimal. But as it stands PT LLMs aren't capable of _exploring_ the joke space, which means they never come out of the RL process with humor. Addendum: Final food for thought. There's kind of this debating going on about "inference scaling", with some believing that CoT, ToT, Reasoning, etc are all essentially just inference scaling. More output gives the model more compute so it can make better predictions. It's likely true that that's the case. In fact, if it _isn't_ the case we need to take a serious look at our training pipelines. But I think it's _also_ about exploring during RL. The extra tokens might give it a boost, sure, but the ability for the model to find more valid paths during RL enables it to express more of its capabilities and solve more problems. If the model is faced with a sheer cliff face it doesn't really matter how much inference compute you throw at it. Only the ability for it to walk around the cliff will help. And, yeah, this all sounds very much like ... gradient descent :P and yes there have been papers on that connection. It very much seems like we're building a second layer of the same stuff here and it's going to be AdamW all the way down.
- dcreater 2y agoWhy ask it for a joke? That's such a bad way to try out a reasoning model
- simonw 2y agoWatching them attempt to reason their way to a joke is really amusing.
- qqqult 2y agoI asked deepseek-14b for a joke in the style of Norm Mcdonald, and it delivered an overexplained deadpan absurd joke that I thought was pretty similar to Norms style. Then I tried again with different styles and realized that all jokes it tries to make are in that same style as if it tries to follow a manual on how to construct a joke without actually being funny. Which is pretty funny
- HeatrayEnjoyer 2y agoIt's a great test. Creating a joke that is both funny and novel requires many steps.
- Weetile 2y agoCreating a joke is a great demonstration of a multistage reasoning/deduction process
- marci 2y agoWould that be funny? Failing a job interview because, when ask to make a joke about a pelican and a walrus who run a tea room together, you couldn't come up with a quip, therefore showing your inability to reason. Let's not even talk about the "r" you forgot when asked to write "cranberry"...
- jonplackett 2y agoThis is probably pretty similar to my inner monologue as I would try and inevitably fail to come up with a good joke.
- laweijfmvo 2y agowhy shouldn’t i assume that the “thinking” is just the usual LLM regurgitation of “how would a human coming up with a joke explain their reasoning?” or something like that, and zero “thinking”?
- brookst 2y agoWhat’s the difference?
- bitwalker 2y agoAt least to me, the difference is that one is ostensibly an explanation of how the AI arrived at the joke, the other is a post-hoc explanation of the joke.
- gpm 2y agoYou can be pretty sure the AI isn't doing a post-hoc explanation because the only writable memory it has access to is the tokens it has already output - i.e. the explanation of the joke. Everything else is reset between every token. As long as it comes up with different jokes different times you ask it (assuming randomness in sampling) - how could it.
- baq 2y agoThe problem is it can’t remember what it hasn’t written but the end result still makes sense, so there has to be some goal after parsing the initial context that the tokens are emitted towards to. This means there’s nothing stopping it from producing an explanation, it might be in there from the very start.
- svachalek 2y agoThere's no goal. The tokens are computed one at a time from all the previous tokens.
- earth2mars 2y agoTried exactly the same model. And unfortunately the reasoning is just useless. Built it is still not able to tell how many r's in strawberry.
- spacemanspiff01 2y agoThat's a tokenizer issue though?
- zamadatix 2y agoNot 100% so for chain of thought models, they should recognize to spell the word letter by letter in some separated form and then count the tokens in that form. The Qwen distill seems to do exactly this really well: > Step-by-step explanation: > 1. Break down each word: "not", "really", "a", "tokenizer", "issue". > 2. Count 'e's in each word: > - "not": 0 > - "really": 1 > - "a": 0 > - "tokenizer": 2 > - "issue": 1 > 3. Sum the counts: 0 + 1 + 0 + 2 + 1 = 4. > > Answer: There are 4 E's in the phrase. In the thought portion it broke the words up every which way you could think to check then validated the total by listing the letters in a number list by index and counting that compared to the sums of when it did each word.
- spacemanspiff01 2y agoBut the only way to do this is if it is trained on how to map the word token to character tokens ie Hello -> h e l l o 66547 -> 12 66 88 88 3 Or, maybe it memorized that hello has a single e. Either way, This seems to be a edge case that may or may not exist in the training data, but seems orthogonal to 'reasoning' A better test case would be how it performs if you give the spelling mappings for each word the context?
- svachalek 2y agoThe important thing is, when we're all replaced by robots, deep down we will know we are superior because we can count letters in strawberry.
- gjm11 2y agoWhat's your sense of how useful local LLMs are for things other than ... writing blog posts about experimenting with local LLMs? :-) (This is a serious question, not poking fun; I am actually curious about this.)
- jhonof 2y agoIf you are worried about security or IP at all, it's preferable to run locally, or spin up your own box that you can query running one of these models.
- gjm11 2y agoI understand the reasons for not wanting to use a remote LLM. My question was about how useful local LLMs are. It might turn out that for some people remote LLMs are unacceptable for privacy reasons and local LLMs are unacceptable because they aren't good enough to be useful.
- simonw 2y agoSix months ago I had almost given up on local LLMs - they were fun to try but they were so much less useful than Sonnet 3.5 / GPT-4o that it was hard to justify using them. That's changed in the past two months. Llama 3 70B, Qwen 32B and now these R1 models are really impressive, to the point that I'm considering trying to get real work done with them. The catch is RAM: I have 64GB, but loading up a current GPT-4 class model uses up around 40GB of that - which doesn't leave much for me to run Firefox and VS Code. So I'm still not likely to use them on a daily basis - but it does make me wonder if I should keep this laptop around as a dedicated server next time I upgrade.
- gjm11 2y agoThanks! One reason why I'm asking is that I'm in the market for a new laptop and am wondering whether it's worth spending more for the possible benefits of being able to run ~30-40GB local LLMs. Unfortunately it doesn't look as if the answer is either "ha ha, obviously not" or "yes, obviously". (If the question were only about models available right now I think the answer would be no, but it seems like they're close enough to being useful that I'm reluctant to bet on them not being clearly genuinely useful a year from now.)
- peeters 2y ago> Wait, maybe the punchline is something like: "We don’t have any fish in the tea, but we do have a lot of krill." Shucks, it was so close to coming up with a good punchline it could work back from. I'm thinking set it in a single-cell comic. A downtrodden young man or woman sitting alone at a table, a pelican in the background clearly making drinks in its voluminous beak, and the walrus waiter places a cup in front of the person, consolingly saying "there's plenty of fish in the tea".
- joenot443 2y agoI chuckled, that's a very clean punchline. Nicely done. These days I sometimes feel as if my skills as an iOS dev are losing market value. Maybe it's time I get back to writing standup :)
- fud101 2y agobut we do have a lot of time to krill would have worked better. I saw it instantly from your post, not sure if it's funny or not though.
- richardw 2y agoWhile not world class, both of those are instantly better than the bag ‘o tokens served by the LLM. Honestly it’s a bit like the porn definition. We know intelligence when we see it, even if it’s hard to define. Adding more tokens doesn’t magically phase-change into real smarts. We’ll get there, but not today.
- patcon 2y agoAnd yet none of it would have ever been experienced without the guidance toward this thinking by an LLM... ;) Maybe future LLMs will be their most pro-social by playing stupid, and letting us feel like we came up with all the good idea they gracelessly-but-gracefully point our attention toward (not saying that's happening here, but maybe it's where we should make it stop)
- scotty79 2y ago
- riwsky 2y ago“I never really had a childhood”, said Walrus, blowing on his tea with a feigned sigh. “Why’s that?” asked Pelican, refilling a sugar shaker. Walrus: “I was born long in the tooth!” Pelican: [big stupid pelican laughing noise]
- fsndz 2y agofrankly ollama + Deepseek is all you need to win with open source AI. I will do some experiments today and add it to my initial blogpost. https://medium.com/thoughts-on-machine-learning/deepseek-is-coming-for-openais-neck-946935f94842 https://medium.com/thoughts-on-machine-learning/deepseek-is-...
- newman314 2y agoHave you had a chance to compare performance and results between the Qwen-7B and Llama-8B versions?