What are the hardware requirements for Tiiuae Falcon 40B Instruct W4 G128 AWQ?

Tiiuae Falcon 40B Instruct W4 G128 AWQ requires approximately 22.3 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Tiiuae Falcon 40B Instruct W4 G128 AWQ and how large is it?

Tiiuae Falcon 40B Instruct W4 G128 AWQ is developed by abhinavkulkarni, a model with 40b 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 Tiiuae Falcon 40B Instruct W4 G128 AWQ?

Tiiuae Falcon 40B Instruct W4 G128 AWQ is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Tiiuae Falcon 40B Instruct W4 G128 AWQ by abhinavkulkarni — VRAM 22.3GB

Name: Tiiuae Falcon 40B Instruct W4 G128 AWQ
Author: abhinavkulkarni

Tiiuae Falcon 40B Instruct W4 G128 AWQ is an open-source language model by abhinavkulkarni. Features: 40b LLM, VRAM: 22.3GB, License: apache-2.0, Quantized, Instruction-Based, LLM Explorer Score: 0.07.

Awq Custom code Instruct Pytorch Quantized Refinedweb Region:us Sharded

Model Card on HF 🤗: https://huggingface.co/abhinavkulkarni/tiiuae-falcon-40b-instruct-w4-g128-awq

Tiiuae Falcon 40B Instruct W4 G128 AWQ Parameters and Internals

Model Type

causal decoder-only, quantized

Additional Notes

The model is finetuned on a mixture of chat/instruct datasets and has been quantized using the AWQ technique.

Supported Languages

English (high proficiency)

Input Output

Accepted Modalities:

text

Output Format:

text

Release Notes

Version:

July 5, 2023

Date:

2023-07-05

Notes:

Initial release of the 4-bit 128 group size AWQ quantized model.

LLM Name	Tiiuae Falcon 40B Instruct W4 G128 AWQ
Repository 🤗	https://huggingface.co/abhinavkulkarni/tiiuae-falcon-40b-instruct-w4-g128-awq
Model Size	40b
Required VRAM	22.3 GB
Updated	2026-08-04
Maintainer	abhinavkulkarni
Model Type	RefinedWeb
Instruction-Based	Yes
Model Files	3.2 GB: 1-of-7 3.2 GB: 2-of-7 3.2 GB: 3-of-7 3.2 GB: 4-of-7 3.2 GB: 5-of-7 3.2 GB: 6-of-7 3.1 GB: 7-of-7
AWQ Quantization	Yes
Quantization Type	awq
Model Architecture	RWForCausalLM
License	apache-2.0
Model Max Length	2048
Transformers Version	4.33.1
Is Biased	0
Tokenizer Class	PreTrainedTokenizerFast
Vocabulary Size	65024
Torch Data Type	float16

Rank the Tiiuae Falcon 40B Instruct W4 G128 AWQ 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 Tiiuae Falcon 40B Instruct W4 G128 AWQ

Best Alternatives	Context / RAM	Downloads	Likes
Falcon 40B Instruct 8bit	0K / 41.8 GB	12	6
Falcon 40B Instruct GPTQ	0K / 22.5 GB	152	197
Falcon 40B Instruct GPTQ	0K / 22.5 GB	11	1
H2ogpt Oig Oasst1 Falcon 40B	0K / 82.5 GB	9	6
...truct GPTQ Inference Endpoints	0K / 22.5 GB	12	2

Note: green Score (e.g. "73.2") means that the model is better than abhinavkulkarni/tiiuae-falcon-40b-instruct-w4-g128-awq.