QuantaSparkLabs/large-ml-dataset
Viewer โข Updated โข 500k โข 765 โข 1
PhysiQuanty PRO
PhysiQuanty
AI & ML interests
Theoretical Physics, Invariant Tokenization, Standard Model of Particle Physics Applied ML ๐ซ๐ท
Recent Activity
liked a dataset about 1 hour ago
QuantaSparkLabs/large-ml-dataset liked a model about 2 hours ago
DaisyChainAI/daisychain-genomics liked a Space about 2 hours ago
DaisyChainAI/Daisychain-Genomics-DemoOrganizations
liked a
dataset about 1 hour ago
liked a
model about 2 hours ago
liked a
Space about 2 hours ago
liked a
dataset about 3 hours ago
liked a
Space about 4 hours ago
liked a
dataset about 4 hours ago
liked a
model about 5 hours ago
reacted to their post with ๐ฅ about 5 hours ago
Post
594
๐ง Arithmetic-SLM : A 30M model that manages to compute simple arithmetic better than a 3B model ๐
WhirlwindAI/Arithmetic-SLM
hugging-apps/arithmetic-slm
๐ Leaderboard ArithMark-2 ๐
๐ฅ Qwen/Qwen2.5-Math-1.5B = 82.08%
๐ฅ WhirlwindAI/Arithmetic-SLM = 78.60% (31.7M Params)
๐ฅ Qwen/Qwen2.5-3B = 78.44%
Example WhirlwindAI/Arithmetic-SLM =
0.5 * 0.5 = 0.25 โ
(132 / 12) + (46 - 15) = 42 โ
(10 + 28) * 3 = 114 โ
1 * (16 + 28) = 44 โ
(21 + 27) * (14 - 7) = 336 โ
By Science AND FOR SCIENCE <3 !
-1/12....
WhirlwindAI/Arithmetic-SLM
hugging-apps/arithmetic-slm
๐ Leaderboard ArithMark-2 ๐
๐ฅ Qwen/Qwen2.5-Math-1.5B = 82.08%
๐ฅ WhirlwindAI/Arithmetic-SLM = 78.60% (31.7M Params)
๐ฅ Qwen/Qwen2.5-3B = 78.44%
Example WhirlwindAI/Arithmetic-SLM =
0.5 * 0.5 = 0.25 โ
(132 / 12) + (46 - 15) = 42 โ
(10 + 28) * 3 = 114 โ
1 * (16 + 28) = 44 โ
(21 + 27) * (14 - 7) = 336 โ
leaderboard = """
| Model | Params | Score |
|----------------------------------|--------------|-----------|
| Qwen/Qwen2.5-Math-1.5B | 1.54B | 82.08% |
| WhirlwindAI/Arithmetic-SLM | 31.70M | 78.60% | <=
| Qwen/Qwen2.5-3B | 3.09B | 78.44% |
| Qwen/Qwen2.5-1.5B | 1.54B | 77.72% |
| Qwen/Qwen2.5-Coder-1.5B | 1.54B | 74.88% |
| HuggingFaceTB/SmolLM2-1.7B | 1.71B | 66.12% |
| Qwen/Qwen2.5-0.5B | 494M | 63.04% |
| facebook/MobileLLM-R1-140M-base | 140M | 53.88% |
| SupraLabs/Supra-50M-Base | 52M | 27.12% |
"""By Science AND FOR SCIENCE <3 !
-1/12....
reacted to their post with ๐ about 6 hours ago
Post
594
๐ง Arithmetic-SLM : A 30M model that manages to compute simple arithmetic better than a 3B model ๐
WhirlwindAI/Arithmetic-SLM
hugging-apps/arithmetic-slm
๐ Leaderboard ArithMark-2 ๐
๐ฅ Qwen/Qwen2.5-Math-1.5B = 82.08%
๐ฅ WhirlwindAI/Arithmetic-SLM = 78.60% (31.7M Params)
๐ฅ Qwen/Qwen2.5-3B = 78.44%
Example WhirlwindAI/Arithmetic-SLM =
0.5 * 0.5 = 0.25 โ
(132 / 12) + (46 - 15) = 42 โ
(10 + 28) * 3 = 114 โ
1 * (16 + 28) = 44 โ
(21 + 27) * (14 - 7) = 336 โ
By Science AND FOR SCIENCE <3 !
-1/12....
WhirlwindAI/Arithmetic-SLM
hugging-apps/arithmetic-slm
๐ Leaderboard ArithMark-2 ๐
๐ฅ Qwen/Qwen2.5-Math-1.5B = 82.08%
๐ฅ WhirlwindAI/Arithmetic-SLM = 78.60% (31.7M Params)
๐ฅ Qwen/Qwen2.5-3B = 78.44%
Example WhirlwindAI/Arithmetic-SLM =
0.5 * 0.5 = 0.25 โ
(132 / 12) + (46 - 15) = 42 โ
(10 + 28) * 3 = 114 โ
1 * (16 + 28) = 44 โ
(21 + 27) * (14 - 7) = 336 โ
leaderboard = """
| Model | Params | Score |
|----------------------------------|--------------|-----------|
| Qwen/Qwen2.5-Math-1.5B | 1.54B | 82.08% |
| WhirlwindAI/Arithmetic-SLM | 31.70M | 78.60% | <=
| Qwen/Qwen2.5-3B | 3.09B | 78.44% |
| Qwen/Qwen2.5-1.5B | 1.54B | 77.72% |
| Qwen/Qwen2.5-Coder-1.5B | 1.54B | 74.88% |
| HuggingFaceTB/SmolLM2-1.7B | 1.71B | 66.12% |
| Qwen/Qwen2.5-0.5B | 494M | 63.04% |
| facebook/MobileLLM-R1-140M-base | 140M | 53.88% |
| SupraLabs/Supra-50M-Base | 52M | 27.12% |
"""By Science AND FOR SCIENCE <3 !
-1/12....
posted an update about 6 hours ago
Post
594
๐ง Arithmetic-SLM : A 30M model that manages to compute simple arithmetic better than a 3B model ๐
WhirlwindAI/Arithmetic-SLM
hugging-apps/arithmetic-slm
๐ Leaderboard ArithMark-2 ๐
๐ฅ Qwen/Qwen2.5-Math-1.5B = 82.08%
๐ฅ WhirlwindAI/Arithmetic-SLM = 78.60% (31.7M Params)
๐ฅ Qwen/Qwen2.5-3B = 78.44%
Example WhirlwindAI/Arithmetic-SLM =
0.5 * 0.5 = 0.25 โ
(132 / 12) + (46 - 15) = 42 โ
(10 + 28) * 3 = 114 โ
1 * (16 + 28) = 44 โ
(21 + 27) * (14 - 7) = 336 โ
By Science AND FOR SCIENCE <3 !
-1/12....
WhirlwindAI/Arithmetic-SLM
hugging-apps/arithmetic-slm
๐ Leaderboard ArithMark-2 ๐
๐ฅ Qwen/Qwen2.5-Math-1.5B = 82.08%
๐ฅ WhirlwindAI/Arithmetic-SLM = 78.60% (31.7M Params)
๐ฅ Qwen/Qwen2.5-3B = 78.44%
Example WhirlwindAI/Arithmetic-SLM =
0.5 * 0.5 = 0.25 โ
(132 / 12) + (46 - 15) = 42 โ
(10 + 28) * 3 = 114 โ
1 * (16 + 28) = 44 โ
(21 + 27) * (14 - 7) = 336 โ
leaderboard = """
| Model | Params | Score |
|----------------------------------|--------------|-----------|
| Qwen/Qwen2.5-Math-1.5B | 1.54B | 82.08% |
| WhirlwindAI/Arithmetic-SLM | 31.70M | 78.60% | <=
| Qwen/Qwen2.5-3B | 3.09B | 78.44% |
| Qwen/Qwen2.5-1.5B | 1.54B | 77.72% |
| Qwen/Qwen2.5-Coder-1.5B | 1.54B | 74.88% |
| HuggingFaceTB/SmolLM2-1.7B | 1.71B | 66.12% |
| Qwen/Qwen2.5-0.5B | 494M | 63.04% |
| facebook/MobileLLM-R1-140M-base | 140M | 53.88% |
| SupraLabs/Supra-50M-Base | 52M | 27.12% |
"""By Science AND FOR SCIENCE <3 !
-1/12....
liked a
model about 11 hours ago
published a
model about 11 hours ago
updated 2
models about 11 hours ago
updated a
model about 14 hours ago
reacted to pranavupadhyaya52's post with ๐ฅ about 15 hours ago
Post
2133
After several weeks of experimenting, debugging, and iterating, I am excited to share WikiSmartBotLM.
WikiSmartBotLM is a compact decoder only language model built from the ground up as an educational and practical project. The goal was not simply to train another language model, but to create one that is easy to understand, modify, and experiment with while following many of the architectural ideas used in modern LLMs.
The model is built on a custom Transformer architecture featuring Rotary Positional Embeddings, RMSNorm, SwiGLU feed forward layers, grouped query attention, and an efficient autoregressive decoder optimized for local inference.
The repository includes the complete model implementation, configuration files, tokenizer integration, training pipeline, inference scripts, checkpoint conversion utilities, and examples that demonstrate how each component works together. Whether you want to understand the forward pass, train your own model, or build applications on top of WikiSmartBotLM, everything is designed to be approachable.
You can directly run the model via the models Huggingfaces space, which I've included in the post.
Model Repository: pranavupadhyaya52/Wiki-SmartBotLM-Instruct
I hope WikiSmartBotLM becomes a useful resource for anyone who enjoys learning by building. Feedback, issues, feature requests, and contributions are always welcome.
https://huggingface.co/spaces/pranavupadhyaya52/WikiSmartBot
WikiSmartBotLM is a compact decoder only language model built from the ground up as an educational and practical project. The goal was not simply to train another language model, but to create one that is easy to understand, modify, and experiment with while following many of the architectural ideas used in modern LLMs.
The model is built on a custom Transformer architecture featuring Rotary Positional Embeddings, RMSNorm, SwiGLU feed forward layers, grouped query attention, and an efficient autoregressive decoder optimized for local inference.
The repository includes the complete model implementation, configuration files, tokenizer integration, training pipeline, inference scripts, checkpoint conversion utilities, and examples that demonstrate how each component works together. Whether you want to understand the forward pass, train your own model, or build applications on top of WikiSmartBotLM, everything is designed to be approachable.
You can directly run the model via the models Huggingfaces space, which I've included in the post.
Model Repository: pranavupadhyaya52/Wiki-SmartBotLM-Instruct
I hope WikiSmartBotLM becomes a useful resource for anyone who enjoys learning by building. Feedback, issues, feature requests, and contributions are always welcome.
https://huggingface.co/spaces/pranavupadhyaya52/WikiSmartBot
- 1 reply
liked a
model 1 day ago
liked a
dataset 1 day ago
upvoted an article 2 days ago
Article
๐ Teaching a 15M French LLM to think deeper โ and to know when to stop ๐ซ๐ท
RDTvlokip
โข โข 3 reacted to RDTvlokip's post with ๐ 2 days ago
Post
1972
I finally changed the architecture of my 15M French LLM. It worked. Then I almost fooled myself about how much and catching that was the real win.
After proving last time that architecture is a threshold, not a lever, I got stubborn: could I change how the model learns? Four honest attempts, Lion, a sharper AdamW ฮฒ2, multi-token prediction, LayerScale. Four failures. The bottleneck wasn't the learning rule either.
So I changed the shape of the computation instead: loop the same transformer blocks 4ร, deeper reasoning, zero added parameters. It beat the baseline on perplexity, the first thing in the whole project to move that number. Then I added my own twist: let each token decide how deep to think, halting on its own entropy.
My first evaluation was spectacular. Coherence up 65%. Hallucinated names down 62%.
It was noise.
Eight prompts, one seed. I re-ran on 50 prompts ร 200 tokens and watched the gains shrink to "modest" and on out-of-domain prompts, recurrence actually made things worse. No universal winner. And none of it is new: it's Adaptive Computation Time (2016), the Universal Transformer (2018), and LoopViT (2026), recombined and measured honestly.
The real lesson:
A number from 8 prompts is a rumor. The eval harness that kills your own best result is worth more than the result it kills. Cite your lineage. Stay preliminary until multiple seeds say otherwise.
The three models are live. The write-up is honest about every caveat ๐
๐ https://huggingface.co/blog/RDTvlokip/teaching-a-15m-french-llm-to-think-deeper
After proving last time that architecture is a threshold, not a lever, I got stubborn: could I change how the model learns? Four honest attempts, Lion, a sharper AdamW ฮฒ2, multi-token prediction, LayerScale. Four failures. The bottleneck wasn't the learning rule either.
So I changed the shape of the computation instead: loop the same transformer blocks 4ร, deeper reasoning, zero added parameters. It beat the baseline on perplexity, the first thing in the whole project to move that number. Then I added my own twist: let each token decide how deep to think, halting on its own entropy.
My first evaluation was spectacular. Coherence up 65%. Hallucinated names down 62%.
It was noise.
Eight prompts, one seed. I re-ran on 50 prompts ร 200 tokens and watched the gains shrink to "modest" and on out-of-domain prompts, recurrence actually made things worse. No universal winner. And none of it is new: it's Adaptive Computation Time (2016), the Universal Transformer (2018), and LoopViT (2026), recombined and measured honestly.
The real lesson:
A number from 8 prompts is a rumor. The eval harness that kills your own best result is worth more than the result it kills. Cite your lineage. Stay preliminary until multiple seeds say otherwise.
The three models are live. The write-up is honest about every caveat ๐
๐ https://huggingface.co/blog/RDTvlokip/teaching-a-15m-french-llm-to-think-deeper
- 6 replies