What are the hardware requirements for Phi 2 GPTQ?

Phi 2 GPTQ requires approximately 1.8 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Phi 2 GPTQ and how large is it?

Phi 2 GPTQ is developed by TheBloke, a model with 2.8b 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 Phi 2 GPTQ?

Phi 2 GPTQ is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Phi 2 GPTQ by TheBloke — VRAM 1.8GB

Name: Phi 2 GPTQ
Author: TheBloke

Phi 2 GPTQ is an open-source language model by TheBloke. Features: 2.8b LLM, VRAM: 1.8GB, License: other, Quantized, LLM Explorer Score: 0.1.

4-bit Base model:microsoft/phi-2 Base model:quantized:microsoft... Code Custom code En Gptq Phi-msft Quantized Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/TheBloke/phi-2-GPTQ

Phi 2 GPTQ Parameters and Internals

Model Type

text-generation, nlp, code

Use Cases

Areas:

research purposes

Applications:

QA format, chat format, code format

Primary Use Cases:

research, exploration of safety challenges

Limitations:

potential for generating inaccurate code and facts, limited code scope to common packages, unreliable instructions adherence, language understanding mostly standard English, potential societal biases

Additional Notes

Intended for research purposes. Not tested for production-level applications.

Supported Languages

en (standard English)

Training Details

Data Sources:

AOAI GPT-3.5, Falcon RefinedWeb, SlimPajama

Data Volume:

1.4T tokens

Methodology:

next-word prediction

Context Length:

2048

Training Time:

14 days

Hardware Used:

96xA100-80G

Model Architecture:

Transformer-based model

Input Output

Accepted Modalities:

text

LLM Name	Phi 2 GPTQ
Repository 🤗	https://huggingface.co/TheBloke/phi-2-GPTQ
Model Name	Phi 2
Model Creator	Microsoft
Base Model(s)	Phi 2 microsoft/phi-2
Model Size	2.8b
Required VRAM	1.8 GB
Updated	2026-07-24
Maintainer	TheBloke
Model Type	phi-msft
Model Files	1.8 GB
Supported Languages	en
GPTQ Quantization	Yes
Quantization Type	gptq
Model Architecture	PhiForCausalLM
License	other
Model Max Length	2048
Transformers Version	4.37.0.dev0
Tokenizer Class	CodeGenTokenizer
Vocabulary Size	51200
Torch Data Type	float16
Activation Function	gelu_new

Rank the Phi 2 GPTQ 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 Phi 2 GPTQ

Best Alternatives	Context / RAM	Downloads	Likes
Phi 2 Quantize Gptq	2K / 1.8 GB	6	2
... 2 Electrical Engineering GPTQ	2K / 1.8 GB	9	7
Phi2 OSST GPTQ	2K / 1.8 GB	9	1
Phi 2 Orange GPTQ	0K / 1.8 GB	17	4
Phi 2 DPO GPTQ	0K / 1.8 GB	37	5
Dolphin 2 6 Phi 2 GPTQ	0K / 1.8 GB	5	2
LMCocktail Phi 2 V1 GPTQ	0K / 1.8 GB	6	1
Bnb DPO 8bit	2K / 3 GB	7	0
Bnb DPO 8bit	2K / 3 GB	7	0
Phi 2 4bit 64rank	2K / 5.6 GB	13	0

Note: green Score (e.g. "73.2") means that the model is better than TheBloke/phi-2-GPTQ.