LLM EXPLORER 67,552 MODELS INDEXED

NanoReason 7B SFT Refined by kowanietz

By kowanietz · 7 downloads

NanoReason 7B SFT Refined is an open-source language model by kowanietz. Features: 7b LLM, VRAM: 14.6GB, Context: 64K, License: apache-2.0, LLM Explorer Score: 0.28.

Base model:finetune:kowanietz/... Base model:kowanietz/nanoreaso...   Conversational Dataset:allenai/dolci-think-sf...   En   Endpoints compatible   Olmo3   Reasoning   Region:us   Safetensors   Sft

NanoReason 7B SFT Refined Parameters and Internals

LLM NameNanoReason 7B SFT Refined
Repository πŸ€—https://huggingface.co/kowanietz/NanoReason-7B-SFT-Refined 
Base Model(s)  NanoReason 7B SFT   kowanietz/NanoReason-7B-SFT
Model Size7b
Required VRAM14.6 GB
Updated2026-11-15
Maintainerkowanietz
Model Typeolmo3
Model Files  14.6 GB
Supported Languagesen
Model ArchitectureOlmo3ForCausalLM
Licenseapache-2.0
Context Length65536
Model Max Length65536
Transformers Version5.17.0
Tokenizer ClassGPT2Tokenizer
Padding Token<|pad|>
Vocabulary Size100278
Torch Data Typebfloat16

Best Alternatives to NanoReason 7B SFT Refined

Best Alternatives
Context / RAM
Downloads
Likes
Olmo 3 7B Instruct64K / 14.6 GB237639159
Olmo 3 7B Instruct SFT64K / 14.6 GB2993576
NanoReason 7B Think64K / 14.6 GB312
Olmo 3 7B Think64K / 14.6 GB5528898
Olmo 3 7B Think Heretic64K / 14.7 GB5861
Olmo 3 1025 7B64K / 14.6 GB12020280
NanoReason 7B SFT64K / 14.6 GB91
TISD OLMo3 7B Instruct64K / 14.6 GB381
Olmo 3 7B Think OPSA64K / 14.6 GB4061
...n Cleaned 10K CPT Best Step75064K / 14.6 GB1450
Note: green Score (e.g. "73.2") means that the model is better than kowanietz/NanoReason-7B-SFT-Refined.