LLM EXPLORER 66,613 MODELS INDEXED

Smol Math Sft Merged by yas1988

By yas1988 · 135 downloads
HF Repo ↗ Copy URL

Smol Math Sft Merged is an open-source language model by yas1988. Features: 1.7b LLM, VRAM: 3.4GB, Context: 8K, License: apache-2.0.

Base model:adapter:huggingface... Base model:huggingfacetb/smoll...   Conversational   Dataset:huggingfacetb/smoltalk   Endpoints compatible   Gsm8k   Llama   Lora   Math   Region:us   Safetensors   Sft   Trl
Model Card on HF 🤗: https://huggingface.co/yas1988/smol-math-sft-merged 

Smol Math Sft Merged Parameters and Internals

LLM NameSmol Math Sft Merged
Repository 🤗https://huggingface.co/yas1988/smol-math-sft-merged 
Base Model(s)  HuggingFaceTB/SmolLM2-1.7B   HuggingFaceTB/SmolLM2-1.7B
Model Size1.7b
Required VRAM3.4 GB
Updated2026-11-04
Maintaineryas1988
Model Typellama
Model Files  3.4 GB
Model ArchitectureLlamaForCausalLM
Licenseapache-2.0
Context Length8192
Model Max Length8192
Transformers Version5.20.2
Tokenizer ClassGPT2Tokenizer
Padding Token<|endoftext|>
Vocabulary Size49152
Errorsreplace

Best Alternatives to Smol Math Sft Merged

Best Alternatives
Context / RAM
Downloads
Likes
...FT Tulu3 Decontaminated Masked16K / 3.4 GB3420
...B 16K SFT Tulu3 Decontaminated16K / 3.4 GB1460
SmolLM2 1.7B 16K16K / 3.4 GB120
SmolLM2 1.7B Instruct 16K16K / 3.4 GB10310
SmolLM2 1.7B Instruct8K / 3.4 GB154013742
SmolLM2 1.7B Instruct8K / 3.4 GB49096
Yoru8K / 3.4 GB3090
Augmented B933f090bb558b888K / 3.4 GB5740
...FT Tulu3 Decontaminated Masked8K / 3.4 GB2740
Banglish Companion8K / 6.8 GB12540
Note: green Score (e.g. "73.2") means that the model is better than yas1988/smol-math-sft-merged.