What are the hardware requirements for Llama3 12B?

Llama3 12B requires approximately 23.1 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 Llama3 12B and how large is it?

Llama3 12B is developed by KingNish, a model with 12b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Llama3 12B perform on standard benchmarks?

Llama3 12B has the following published scores: MMLU 66.47. Compare against reference models on this page or on the LLM Explorer leaderboards.

Llama3 12B by KingNish — VRAM 23.1GB, 8K context

Name: Llama3 12B
Author: KingNish

Llama3 12B is an open-source language model by KingNish. Features: 12b LLM, VRAM: 23.1GB, Context: 8K, License: llama3, Merged, LLM Explorer Score: 0.14, Arc: 57.5, HellaSwag: 75.9, MMLU: 66.5, GSM8K: 33.1.

Merged Model Abhishek/autotrain-llama3-orpo... Autotrain compatible Base model:abhishek/autotrain-... Base model:finetune:abhishek/a... Conversational Endpoints compatible Llama Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/KingNish/Llama3-12b

Llama3 12B Parameters and Internals

Model Type

text-generation

Additional Notes

Model is a merge of abhishek/autotrain-llama3-orpo-v2 using LazyMergekit.

LLM Name	Llama3 12B
Repository 🤗	https://huggingface.co/KingNish/Llama3-12b
Base Model(s)	Autotrain Llama3 Orpo V2 Autotrain Llama3 Orpo V2 abhishek/autotrain-llama3-orpo-v2 abhishek/autotrain-llama3-orpo-v2
Merged Model	Yes
Model Size	12b
Required VRAM	23.1 GB
Updated	2024-12-10
Maintainer	KingNish
Model Type	llama
Model Files	1.1 GB: 1-of-13 1.9 GB: 2-of-13 2.0 GB: 3-of-13 2.0 GB: 4-of-13 2.0 GB: 5-of-13 1.9 GB: 6-of-13 1.9 GB: 7-of-13 2.0 GB: 8-of-13 1.9 GB: 9-of-13 2.0 GB: 10-of-13 1.9 GB: 11-of-13 2.0 GB: 12-of-13 0.5 GB: 13-of-13
Model Architecture	LlamaForCausalLM
License	llama3
Context Length	8192
Model Max Length	8192
Transformers Version	4.40.1
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<pad>
Vocabulary Size	128264
Torch Data Type	bfloat16

Rank the Llama3 12B 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

Best Alternatives to Llama3 12B

Best Alternatives	Context / RAM	Downloads	Likes
L3.1 Smth.Concv3 12B	128K / 47.8 GB	56	0
L3.1 Legionv2 12B	128K / 47.8 GB	20	0
L3.1 Nimbusv1 12B	128K / 48.1 GB	5	0
L3.1 Adumbralv1 12B	128K / 47.8 GB	6	0
L3.1 Punicav1 12B	128K / 48.1 GB	12	0
L3.1 Twilightv1 12B	128K / 48.1 GB	6	0
OpenCrystal 12B L3.1 128K	128K / 23 GB	4	3
OpenCrystal 12B L3	8K / 23 GB	17	17
IxChel L3 12B	8K / 23 GB	13	2
Ursidae 12B Mini	8K / 23 GB	20	3

Note: green Score (e.g. "73.2") means that the model is better than KingNish/Llama3-12b.