What are the hardware requirements for PULI LlumiX 32K Instruct F16 0.2?

PULI LlumiX 32K Instruct F16 0.2 requires approximately 13.5 GB of VRAM and supports a context window of 32K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed PULI LlumiX 32K Instruct F16 0.2 and how large is it?

PULI LlumiX 32K Instruct F16 0.2 is developed by ariel-ml, a model with 6.7b 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 PULI LlumiX 32K Instruct F16 0.2?

PULI LlumiX 32K Instruct F16 0.2 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

PULI LlumiX 32K Instruct F16 0.2 by ariel-ml — VRAM 13.5GB, 32K context

Name: PULI LlumiX 32K Instruct F16 0.2
Author: ariel-ml

PULI LlumiX 32K Instruct F16 0.2 is an open-source language model by ariel-ml. Features: 6.7b LLM, VRAM: 13.5GB, Context: 32K, License: llama2, Quantized, Fine-Tuned, Instruction-Based, LLM Explorer Score: 0.12.

Base model:finetune:nytk/puli-... Base model:nytk/puli-llumix-32... Conversational Custom code Dataset:boapps/szurkemarha En Endpoints compatible Finetuned Fp16 Hu Instruct Llama Puli Quantized Region:us Safetensors Sharded Tensorflow Trl Unsloth

Model Card on HF 🤗: https://huggingface.co/ariel-ml/PULI-LlumiX-32K-instruct-f16-0.2

PULI LlumiX 32K Instruct F16 0.2 Parameters and Internals

Model Type

text-generation

Use Cases

Areas:

research, commercial applications

Limitations:

max_seq_length = 32 768, float16, vocab size: 32 000

Supported Languages

hu (High), en (Moderate)

Training Details

Data Sources:

boapps/szurkemarha

Data Volume:

~53000 prompts

Methodology:

The model was finetuned with LoRA techniques using parameters like LoRA rank of 16 and LoRA alpha of 16. Continuous pretraining on a Hungarian dataset was also done.

Context Length:

32768

Hardware Used:

RTX 4090 GPU

LLM Name	PULI LlumiX 32K Instruct F16 0.2
Repository 🤗	https://huggingface.co/ariel-ml/PULI-LlumiX-32K-instruct-f16-0.2
Base Model(s)	PULI LlumiX 32K NYTK/PULI-LlumiX-32K
Model Size	6.7b
Required VRAM	13.5 GB
Updated	2026-08-01
Maintainer	ariel-ml
Model Type	llama
Instruction-Based	Yes
Model Files	4.9 GB: 1-of-3 5.0 GB: 2-of-3 3.6 GB: 3-of-3
Supported Languages	hu en
Quantization Type	fp16
Model Architecture	LlamaForCausalLM
License	llama2
Context Length	32768
Model Max Length	32768
Transformers Version	4.40.2
Tokenizer Class	LlamaTokenizer
Padding Token	<unk>
Vocabulary Size	32000
Torch Data Type	bfloat16

Rank the PULI LlumiX 32K Instruct F16 0.2 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 PULI LlumiX 32K Instruct F16 0.2

Best Alternatives	Context / RAM	Downloads	Likes
...LI LlumiX 32K Instruct F16 0.1	32K / 13.5 GB	10	0
...rpreter DS 6.7B 6.0bpw H6 EXL2	16K / 5.2 GB	5	2
...rpreter DS 6.7B 8.0bpw H8 EXL2	16K / 6.9 GB	3	2
...rpreter DS 6.7B 4.0bpw H6 EXL2	16K / 3.6 GB	1	1
...coder S DS 6.7B 4.0bpw H6 EXL2	16K / 3.6 GB	3	5
...coder S DS 6.7B 3.0bpw H6 EXL2	16K / 2.8 GB	6	3
...coder S DS 6.7B 6.0bpw H6 EXL2	16K / 5.2 GB	5	3
...coder S DS 6.7B 5.0bpw H6 EXL2	16K / 4.4 GB	7	2
...coder S DS 6.7B 8.0bpw H8 EXL2	16K / 6.9 GB	2	1
...epseek Coder 6.7B Instruct AWQ	16K / 3.9 GB	291702	19

Note: green Score (e.g. "73.2") means that the model is better than ariel-ml/PULI-LlumiX-32K-instruct-f16-0.2.