What are the hardware requirements for Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed?

Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed requires approximately 1.3 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.

Where can I download or evaluate Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed?

Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed by PrunaAI — VRAM 1.3GB, 8K context

Name: Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed
Author: PrunaAI

Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed is an open-source language model by PrunaAI. Features: 2b LLM, VRAM: 1.3GB, Context: 8K, Quantized, LLM Explorer Score: 0.13.

1bit Base model:cognitivecomputatio... Base model:finetune:cognitivec... Gemma Pruna-ai Quantized Region:us

Model Card on HF 🤗: https://huggingface.co/PrunaAI/cognitivecomputations-dolphin-2.8-gemma-2b-HQQ-1bit-smashed

Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed Parameters and Internals

Model Type

compressed

Use Cases

Areas:

model compression, efficiency improvement

Primary Use Cases:

reducing AI model size, improving model inference speed, decreasing energy consumption

Additional Notes

Model efficiency and metrics provided for first run and sync/async distinction. Compressed using 1bit quantization.

Training Details

Data Sources:

WikiText

Methodology:

Model is compressed with hqq.

Input Output

Input Format:

text

Accepted Modalities:

text

Output Format:

text

Performance Tips:

Run efficiency tests under intended use-case conditions.

LLM Name	Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed
Repository 🤗	https://huggingface.co/PrunaAI/cognitivecomputations-dolphin-2.8-gemma-2b-HQQ-1bit-smashed
Base Model(s)	cognitivecomputations/dolphin-2.8-gemma-2b cognitivecomputations/dolphin-2.8-gemma-2b
Model Size	2b
Required VRAM	1.3 GB
Updated	2025-09-13
Maintainer	PrunaAI
Model Type	gemma
Model Files	1.3 GB
Quantization Type	1bit
Model Architecture	GemmaForCausalLM
Context Length	8192
Model Max Length	8192
Transformers Version	4.40.0
Tokenizer Class	GemmaTokenizer
Padding Token	<pad>
Vocabulary Size	256000
Torch Data Type	float16

Rank the Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed 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 Cognitivecomputations Dolphin 2.8 Gemma 2B HQQ 1bit Smashed

Best Alternatives	Context / RAM	Downloads	Likes
... 2B It Hermes Function Calling	8K / 5.1 GB	18	0
Vi Gemma 2B RAG	8K / 5.1 GB	36	13
My AwesomeFinance Model	8K / 2.1 GB	7	0
... 2.8 Gemma 2B HQQ 2bit Smashed	8K / 1.6 GB	5	0
... 2.8 Gemma 2B HQQ 4bit Smashed	8K / 2.1 GB	5	0
Emo AI 3B	8K / 5.1 GB	6	1
Jenna V3 QGemma 2B	8K / 5.1 GB	6	0
Gemma 2B Sentiment	8K / 5.1 GB	13	0
Jenna V2	8K / 5.1 GB	6	0
... 2.8 Gemma 2B Bnb 4bit Smashed	8K / 2.2 GB	5	0

Note: green Score (e.g. "73.2") means that the model is better than PrunaAI/cognitivecomputations-dolphin-2.8-gemma-2b-HQQ-1bit-smashed.