What are the hardware requirements for SmolLM2 360M Instruct?

SmolLM2 360M Instruct requires approximately 0.7 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed SmolLM2 360M Instruct and how large is it?

SmolLM2 360M Instruct is developed by HuggingFaceTB, a model with 360m parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate SmolLM2 360M Instruct?

SmolLM2 360M Instruct is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

SmolLM2 360M Instruct by HuggingFaceTB — VRAM 0.7GB, 8K context

Name: SmolLM2 360M Instruct
Author: HuggingFaceTB

SmolLM2 360M Instruct is an open-source language model by HuggingFaceTB. Features: 360m LLM, VRAM: 0.7GB, Context: 8K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.27.

Arxiv:2502.02737 Base model:huggingfacetb/smoll... Base model:quantized:huggingfa... Conversational En Endpoints compatible Instruct Llama Onnx Region:us Safetensors Tensorboard Transformers.js

Model Card on HF 🤗: https://huggingface.co/HuggingFaceTB/SmolLM2-360M-Instruct

SmolLM2 360M Instruct Parameters and Internals

Model Type

text generation

Use Cases

Areas:

Research, Commercial applications

Limitations:

Primarily understands and generates English, Generated content may not always be factually accurate, May not be logically consistent or free from biases

Considerations:

Should be used as assistive tools; verify important information.

Supported Languages

en (High)

Training Details

Data Sources:

FineWeb-Edu, DCLM, The Stack, Custom curated datasets

Data Volume:

4T tokens

Methodology:

Supervised fine-tuning with Direct Preference Optimization using UltraFeedback

Hardware Used:

64 H100 GPUs

Model Architecture:

Transformer decoder

Input Output

Input Format:

Input chat template through tokenizer

Accepted Modalities:

text

Output Format:

Generated text responses

Performance Tips:

Utilize GPUs for optimal performance.

LLM Name	SmolLM2 360M Instruct
Repository 🤗	https://huggingface.co/HuggingFaceTB/SmolLM2-360M-Instruct
Base Model(s)	SmolLM2 360M HuggingFaceTB/SmolLM2-360M
Model Size	360m
Required VRAM	0.7 GB
Updated	2026-08-08
Maintainer	HuggingFaceTB
Model Type	llama
Instruction-Based	Yes
Model Files	0.7 GB 0.0 GB
Supported Languages	en
Context Length	8k
Model Architecture	LlamaForCausalLM
License	apache-2.0
Context Length	8192
Model Max Length	8192
Transformers Version	4.42.3
Tokenizer Class	GPT2Tokenizer
Padding Token	<\|im_end\|>
Vocabulary Size	49152
Torch Data Type	bfloat16

Rank the SmolLM2 360M Instruct capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Quantized Models of the SmolLM2 360M Instruct

Model	Likes	Downloads
Deep Thought Posttrain	1	782
SmolLM2 360M Instruct Bnb 4bit	3	520
SmolLM2 360M Instruct Q8 Mlx	1	93
SmolLM2 360M Bnb 4bit	2	46

Best Alternatives to SmolLM2 360M Instruct

Best Alternatives	Context / RAM	Downloads	Likes
Smol Lm2 360 Instruct	8K / 0.7 GB	313	3
Smollm2 360M Text2sql Sft	8K / 1.4 GB	194	0
Smollm3 720prms	8K / 0.7 GB	5	0
ProseFlow V1 360M Instruct	8K / 0.7 GB	5	1
SmolLM2 Rethink 360M	8K / 1.4 GB	13	1
SolaraV2 Coder 0511	8K / 0.7 GB	13	1
BrainrotLM Assistant 362M	8K / 0 GB	10	0
SmolLM2 CoT 360M	8K / 1.4 GB	62	9
...n Combined Instruction Dataset	8K / 1.4 GB	6	1
... Cpt Fineweb Norwegian Nynorsk	8K / 1.4 GB	9	0