9 ms·
Smollm3: Smol, multilingual, long-context reasoner LLM
- gardnr 1y agoIt's small (3B) and does great on benchmarks. This is a model for edge / mobile deployments so the gains over gemma3-4b are meaningful. It has dual mode reasoning / non_reasoning AND they released the full training method: > We're releasing SmolLM3 with our engineering blueprint. It includes architecture details, exact data mixtures showing how we progressively boost performance across domains in a three-stage pretraining approach, and the methodology for building a hybrid reasoning model. Usually, achieving these results would require months of reverse engineering. Instead, we're providing the full methodology.
- sigmoid10 1y agoI hate to say it, but reasoning models simply aren't suited for edge computing. I just ran some tests on this model and even at 4bit weight quantisation it blows past 10GB of VRAM with just ~1000 tokens while it is still reasoning. So even if you're running on a dedicated ML edge device like a $250 Jetson, you will run out of memory before the model even formulates a real answer. You'll need a high end GPU to make full use of it for limited answers and an enterprise grade system to support longer contexts. And with reasoning turned off I don't see any meaningful improvement over older models. So this is primarily great for enterprises who want to do on-prem with limited budgets and maybe high-end enthusiasts.
- wizee 1y agoYou should use flash attention with KV cache quantization. I routinely use Qwen 3 14B with the full 128k context and it fits in under 24 GB VRAM. On my Pixel 8, I've successfully used Qwen 3 4B with 8K context (again with flash attention and KV cache quantization).
- sigmoid10 1y ago>On my Pixel 8, I've successfully used Qwen 3 4B How many tokens/s? I can't imagine that this would run in any practical way.
- tiahura 1y agoCan anyone estimate how much of the 3B is necessitated by multi-language support?
- rockinghigh 1y agoThe vocabulary size is fairly small (128,256) for a multilingual model. I would guess it doesn't require many additional parameters to support these 5 languages as many tokens can be shared.
- ethan_smith 1y agoTypically, multilingual capabilities consume 20-30% of model parameters in small LLMs, primarily in token embeddings and early transformer layers. Monolingual variants of similar models often perform better on English benchmarks with the same parameter count.
- netdur 1y agonaive look, 2/3 of model, without multi-languages this shiuld be around 1B
- nateb2022 1y agohttps://web.archive.org/web/20250708164705/https://huggingface.co/blog/smollm3 https://web.archive.org/web/20250708164705/https://huggingfa...
- _1 1y agoWhich small model is good for fine tuning to various enterprise data sets? Our business units are wanting to run small models in browser and on mobile devices, without dealing with RAG and cloud resources.
- mhitza 1y agoYou really need to try them all out yourself and make sure you have proper benchmarks. While machine learning is not my field, I've tried to finetune Mistral 7B (following their official guide and toolset) and the results did not satisfy. Had a few very specific questions from the dataset that no matter how much I've finetuned and tweaked the process it was not able to respond with correct information. A mix of vector search + keyword search is still better at building the right question context than expecting it to learn all the information. I've used the pretrained dataset approach. Maybe building syntethic questions and answers around the dataset yields better results but I didn't have time to experiment with that approach.
- ivape 1y agoHow much data did you use to fine tune?
- mhitza 1y agoKilobytes to megabytes of data. I was trying to fine-tune it for some specific legislation I was expecting to be able afterwards to ask about.
- magicalhippo 1y ago> Maybe building syntethic questions and answers around the dataset yields better results but I didn't have time to experiment with that approach. While they answer a slightly different question in the Physics of Language Models[1], based on their results it seems to me it is likely that one needs to do such augmentation of the dataset to get good results. However, they also show that the dataset the base model is trained on can drastically affect finetuning performance. So if the base model is trained on a poor dataset for your specific task, perhaps you'll never get good performance. [1]: https://physics.allen-zhu.com/part-3-knowledge/part-3-1 https://physics.allen-zhu.com/part-3-knowledge/part-3-1
- WhitneyLand 1y agoMostly SOTA performance at the 3B level. A notable addition to the small but truly open club of models that provide full disclosure, code, recipes to reproduce their work. Looks like ballpark a million dollars of GPU time if you want to train up one for yourself (4000 gpus/24 days). Very nice write up that’s generous in sharing their learnings. This is a solid and positive contribution.
- YetAnotherNick 1y agoIt's 384 H100s for 24 days, costing less than half a million dollars.
- Imustaskforhelp 1y agoPardon me, but is the dataset public. Like if I really really just wanted to build it from scratch, could I do so? (not that I have that money but just curious)
- hynky 1y agoyes, both core web datasets are publicly available as well as the rest
- Imustaskforhelp 1y agoThanks! To be honest, if I might argue then that this is one of the best truly open source models that we have got. There is AllenAI and (Elmo?) and there is also this one which does distributed training but I think this looks a lot like SOTA for 3B parameters to me. Thanks for telling me, I am not going to lie, I am going to try to test it now! (Ima try some GGUF since ollama convenience)
- peatmoss 1y agoOLMo: https://allenai.org/olmo https://allenai.org/olmo AFAIK, they were the first open everything model.
- bitwize 1y agoThere's a British comedy skit lurking in here. "So it's a small large language model?" "Oh yes, very small." "How can it be small and large at the same time?" "Well, it's small by the standards of a large language model." "So it's large." "Oh yes, very large." "Large compared to what?" "Small language models." "And so something like ChatGPT, what would that be exactly? A large large language model?" "Yes, precisely. An LLLM."
- netdur 1y agoit's big little planet or small big planet?
- janalsncm 1y agoStandards have shifted as well. Gpt2 used to be considered “large” but it is half the size of this. Oh and also Sam Altman said it was too dangerous to release. At this point I consider anything too big to run on consumer grade hardware to be large, but an exact definition is a little silly to argue about.
- a_wild_dandan 1y agoAltman released GPT-2 despite expressing that doing so was a bad idea? That's wild.
- Alifatisk 1y agoI think Altman meant it's too dangerous to open-source GPT-2, therefore locked it in behind a service.
- janalsncm 1y agoIt’s not locked behind a service though. https://huggingface.co/openai-community/gpt2/blob/main/model.safetensors https://huggingface.co/openai-community/gpt2/blob/main/model...
- 1y ago
- msgodel 1y agoWow. Close to a Qwen3 distill with 75% the size. That's great! I've been using the smollm base models for my own finetunes just because they're so high quality, it looks like I might be using them to drive local agents/code completion in the near future too. Their RL algorithm looks interesting. I'm still using OpenAI's algorithm for my stuff, I've been meaning to check on the SoTA since I know my code is pretty outdated (It's crazy how fast that happens with this stuff.)
- gdiamos 1y agoNice work anton et al. I hope you continue the 50-100M parameter models. I think there is a case for models that finish fast on CPUs in solve by llm test cases.
- eachro 1y agoFrom what I've heard, the llama3 models are fairly easy to fine-tune (please correct me if I'm wrong or if there are more amenable models here). How easy is it to finetune smollm3? I know a lot of the MoE LLMs have been quite fickle in this regard.
- BarakWidawsky 1y agoIt’s interesting that it looks like they didn’t apply their own RL to the model, and instead fine tuned on reasoning traces from large datasets and generating reasoning traces from larger models
- lewtun 1y agoIndeed we opted for offline methods like Anchored Preference Optimization as we found in the Open R1 project that doing multi-task RL on small models is quite a hassle to get right. With offline methods, you focus much more on dataset curation / generation, but that still provides faster iteration cycles for the model scale we’re dealing with!
- ivape 1y agoI wonder if this will be cheaper than llama 3.1 8b on OpenRouter.
- ivape 1y agoLooks like it's the 3B models that are being shipped out to on device by default. Apple's on-device LLM is 3B, and I believe Canary is shipping Google nano: https://developer.chrome.com/docs/ai/rewriter-api https://developer.chrome.com/docs/ai/rewriter-api
- danielhanchen 1y agoI fixed some chat template issues for llama.cpp and other inference engines! To run it, do: ./llama.cpp/llama-cli -hf unsloth/SmolLM3-3B-GGUF:Q4_K_XL --jinja -ngl 99
- segmondy 1y agodoing the good work, thanks daniel!
- danielhanchen 1y agoThank you!
- v5v3 1y agoThanks
- danielhanchen 1y agoThanks!
- diggan 1y ago> fixed some chat template issues This seems to be a persistent issue with almost all weight releases, even from bigger companies like Meta. Are the people who release these weights not testing them in various inference engines? Seems they make it work with Huggingface's Transformers library, then call it a day, but sometimes not even that.
- clarionbell 1y agoNo they don't. Why would they? Most of them are using a single inference engine, most likely developed inhouse. Or they go for something like vLLM, but llama.cpp especially is under their radar. The reason is simple. There isn't much money in it. llama.cpp is free and targets lower end of the hardware spectrum. Corporations will run something else, or even more likely, offload the task to contractor.
- simonw 1y agoI'm having trouble running this on my Mac - I've tried Ollama and llama.cpp llama-server so far, both using GGUFs from Hugging Face, but neither worked. (llama_model_load: error loading model: error loading model architecture: unknown model architecture: 'smollm3') I've managed to run it using Python and transformers with PyTorch in device="cpu" mode but unsurprisingly that's really slow - it took 35s to respond to "say hi"! Anyone had success with this on a Mac yet? I really want to get this running with tool calling, ideally via an OpenAI-compatible serving layer like llama-server.
- tripplyons 1y agoHave you tried setting device="mps" to use Metal? It should be faster than PyTorch's "cpu" device on Mac.
- reach-vb 1y agoHey Simon, VB from Hugging Face here and the person who added the model to MLX and llama.cpp (with Son). The PR hasn't yet landed on llama.cpp, hence it doesn't work OTB on llama.cpp installed via brew (similarly doesn't work with ollama since they need to bump their llama.cpp runtime) The easiest would be to install llama.cpp from source: https://github.com/ggml-org/llama.cpp https://github.com/ggml-org/llama.cpp If you want to avoid it, I added SmolLM3 to MLX-LM as well: You can run it via `mlx_lm.chat --model "mlx-community/SmolLM3-3B-bf16"` (requires the latest mlx-lm to be installed) here's the MLX-lm PR if you're interested: https://github.com/ml-explore/mlx-lm/pull/272 https://github.com/ml-explore/mlx-lm/pull/272 similarly, llama.cpp here: https://github.com/ggml-org/llama.cpp/pull/14581 https://github.com/ggml-org/llama.cpp/pull/14581 Let me know if you face any issues!
- kosolam 1y agoCould you please enlighten me regarding all these engines, I’m using lamacpp and ollama. Should I also try mlx, onnx, vllm, etc. I’m not quite sure whats the difference between all these. I’m running on CPU and sometimes GPU
- 1y ago
- grrowl 1y agoGreat to see Huggingface stick to their guns with CodeEval and python tooling. Agentic turn-by-turn tool calling is fine and all, but we're underutilising their ability to write an execute code in an "agent-like" environment.
- cess11 1y agoI've tried to use gemma3:4b which comes up better in that benchmark and found it to be quite disappointing. It breaks a lot, sucks even worse than qwen2.5-coder:7b and incept5/llama3.1-claude:7b at code, needs to be tricked or threatened into saying stuff about many everyday topics. It also commonly chugs away for minutes exercising the GPU fans before responding, at which point I'm already ahead because I figured out another way to solve my problem or get at some information. My experience with phi4-mini and granite3.3 was about the same, and they annoy me even more when I hook them into code editors and try to get them to contribute to my work. For one because they're slow, and at best they suggest adding unnecessary error handling in the style of null checks everywhere, at worst they just start mixing or hallucinating programming languages. Where they would be useful as leverage if they worked, i.e. close to the edge of where I can debug and refactor without getting stuck, they just go into straight nonsense mode, especially on terse first-pass code. Sometimes I've tried to query these things for descriptions of recent history in foreign countries, Wikipedia trivia basically, and they're very often wrong in subtle ways. For example, a politician might have been at it for half a century or so in a troubled country and because they've been ousted in a coup once in the eighties the model is absolutely sure they can't have been in office since. If a person acted like these things do I'd wish for them to get immediate institutional care. Maybe the problem is somehow with me, but I have a deep suspicion it's not.
- iamnotagenius 1y ago[dead]
- lvl155 1y agoThis is actually good learning material for anyone getting up to speed on LLM from scratch.