What are the hardware requirements for RAFT Llama3 Version 24.06.01?

RAFT Llama3 Version 24.06.01 requires approximately 5.8 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 RAFT Llama3 Version 24.06.01 and how large is it?

RAFT Llama3 Version 24.06.01 is developed by Vipinap, a model with 8.2b 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 RAFT Llama3 Version 24.06.01?

RAFT Llama3 Version 24.06.01 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

RAFT Llama3 Version 24.06.01 by Vipinap — VRAM 5.8GB, 8K context

Name: RAFT Llama3 Version 24.06.01
Author: Vipinap

RAFT Llama3 Version 24.06.01 is an open-source language model by Vipinap. Features: 8.2b LLM, VRAM: 5.8GB, Context: 8K, License: apache-2.0, Merged, LLM Explorer Score: 0.13.

Merged Model 4-bit Bitsandbytes Conversational En Endpoints compatible Llama Region:us Safetensors Sharded Tensorflow Trl Unsloth

Model Card on HF 🤗: https://huggingface.co/Vipinap/RAFT_llama3_version_24.06.01

RAFT Llama3 Version 24.06.01 Parameters and Internals

Model Type

text-generation-inference, transformers

Additional Notes

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.

LLM Name	RAFT Llama3 Version 24.06.01
Repository 🤗	https://huggingface.co/Vipinap/RAFT_llama3_version_24.06.01
Base Model(s)	RAFT_llama3_version_Merged_24.06.01 /RAFT_llama3_version_Merged_24.06.01
Merged Model	Yes
Model Size	8.2b
Required VRAM	5.8 GB
Updated	2026-05-23
Maintainer	Vipinap
Model Type	llama
Model Files	4.7 GB: 1-of-2 1.1 GB: 2-of-2
Supported Languages	en
Model Architecture	LlamaForCausalLM
License	apache-2.0
Context Length	8192
Model Max Length	8192
Transformers Version	4.39.3
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|reserved_special_token_250\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the RAFT Llama3 Version 24.06.01 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 RAFT Llama3 Version 24.06.01

Best Alternatives	Context / RAM	Downloads	Likes
Ko Pt Model Test1	8K / 16.4 GB	2027	0
Hola	8K / 5.8 GB	6	0
Llama SciQ 4bits	8K / 5.8 GB	12	0
...ma3 Quantized 5epochs 24 06 04	8K / 5.8 GB	7	0
Llama 3 Merged Linear	8K / 5.8 GB	16	0
Llama 3 Ko Luxia Instruct	8K / 16.4 GB	6	3
KernelLLM Unsloth Bnb 4bit	128K / 6 GB	6	2
KernelLLM Bnb 4bit	128K / 5.7 GB	9	1
Llama3.1 Merged	128K / 5.8 GB	6	0
EPFL TA Meister 4bit	8K / 5.8 GB	6	0

Note: green Score (e.g. "73.2") means that the model is better than Vipinap/RAFT_llama3_version_24.06.01.