What are the hardware requirements for Poro 34B?

Poro 34B requires approximately 68 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Poro 34B and how large is it?

Poro 34B is developed by LumiOpen, a model with 34b 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 Poro 34B?

Poro 34B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Poro 34B by LumiOpen — VRAM 68GB

Name: Poro 34B
Author: LumiOpen

Poro 34B is an open-source language model by LumiOpen. Features: 34b LLM, VRAM: 68GB, License: apache-2.0, LLM Explorer Score: 0.1.

Arxiv:2404.01856 Bloom Dataset:allenai/dolma Dataset:bigcode/starcoderdata Dataset:cerebras/slimpajama-62... Dataset:mc4 Deploy:azure En Endpoints compatible Fi Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/LumiOpen/Poro-34B

Poro 34B Parameters and Internals

Model Type

text generation

Additional Notes

Poro is an advanced language model optimized primarily for English, Finnish, and code.

Supported Languages

fi (high), en (high)

Training Details

Data Sources:

cerebras/SlimPajama-627B, bigcode/starcoderdata, mc4, allenai/dolma

Data Volume:

1 trillion tokens

Methodology:

BLOOM architecture with ALiBi embeddings

Context Length:

2048

Hardware Used:

LUMI supercomputer, 512 AMD MI250X GPUs

Model Architecture:

decoder-only transformer

Input Output

Accepted Modalities:

text

LLM Name	Poro 34B
Repository 🤗	https://huggingface.co/LumiOpen/Poro-34B
Model Size	34b
Required VRAM	68 GB
Updated	2026-08-08
Maintainer	LumiOpen
Model Type	bloom
Model Files	4.7 GB: 1-of-14 4.9 GB: 2-of-14 4.9 GB: 3-of-14 4.9 GB: 4-of-14 4.9 GB: 5-of-14 4.9 GB: 6-of-14 4.9 GB: 7-of-14 4.9 GB: 8-of-14 4.9 GB: 9-of-14 4.9 GB: 10-of-14 4.9 GB: 11-of-14 4.9 GB: 12-of-14 4.9 GB: 13-of-14 4.5 GB: 14-of-14
Supported Languages	fi en
Model Architecture	BloomForCausalLM
License	apache-2.0
Transformers Version	4.37.2
Tokenizer Class	BloomTokenizer
Padding Token	<pad>
Vocabulary Size	128000
Torch Data Type	bfloat16

Rank the Poro 34B 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 Poro 34B

Model	Likes	Downloads	VRAM
Poro 34B GPTQ	3	11	20 GB
Poro 34B GGUF	5	623	14 GB
Poro 34B AWQ	2	7	21 GB

Best Alternatives to Poro 34B

Best Alternatives	Context / RAM	Downloads	Likes
Poro 34B Chat	0K / 68 GB	83	13
Poro 34B GPTQ	0K / 20.3 GB	11	3
Poro 34B AWQ	0K / 21 GB	7	2
Poro 34B AWQ	0K / 21 GB	6	1

Note: green Score (e.g. "73.2") means that the model is better than LumiOpen/Poro-34B.