You can deploy this model and start earning money today!

RichardErkhov/
oopere_-_pruned60-llama-3.2-3b-gguf

Name: oopere_-_pruned60-llama-3.2-3b-gguf
Availability: InStock

endpoints_compatible

gguf

template

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

pruned60-llama-3.2-3b - GGUF

Model creator: https://huggingface.co/oopere/
Original model: https://huggingface.co/oopere/pruned60-llama-3.2-3b/

Name	Quant method	Size
pruned60-llama-3.2-3b.Q2_K.gguf	Q2_K	1.25GB
pruned60-llama-3.2-3b.IQ3_XS.gguf	IQ3_XS	1.31GB
pruned60-llama-3.2-3b.IQ3_S.gguf	IQ3_S	1.34GB
pruned60-llama-3.2-3b.Q3_K_S.gguf	Q3_K_S	1.34GB
pruned60-llama-3.2-3b.IQ3_M.gguf	IQ3_M	1.39GB
pruned60-llama-3.2-3b.Q3_K.gguf	Q3_K	1.39GB
pruned60-llama-3.2-3b.Q3_K_M.gguf	Q3_K_M	1.39GB
pruned60-llama-3.2-3b.Q3_K_L.gguf	Q3_K_L	1.43GB
pruned60-llama-3.2-3b.IQ4_XS.gguf	IQ4_XS	1.46GB
pruned60-llama-3.2-3b.Q4_0.gguf	Q4_0	0.32GB
pruned60-llama-3.2-3b.IQ4_NL.gguf	IQ4_NL	0.45GB
pruned60-llama-3.2-3b.Q4_K_S.gguf	Q4_K_S	1.5GB
pruned60-llama-3.2-3b.Q4_K.gguf	Q4_K	1.51GB
pruned60-llama-3.2-3b.Q4_K_M.gguf	Q4_K_M	1.51GB
pruned60-llama-3.2-3b.Q4_1.gguf	Q4_1	0.32GB
pruned60-llama-3.2-3b.Q5_0.gguf	Q5_0	0.32GB
pruned60-llama-3.2-3b.Q5_K_S.gguf	Q5_K_S	1.65GB
pruned60-llama-3.2-3b.Q5_K.gguf	Q5_K	1.65GB
pruned60-llama-3.2-3b.Q5_K_M.gguf	Q5_K_M	1.65GB
pruned60-llama-3.2-3b.Q5_1.gguf	Q5_1	0.33GB
pruned60-llama-3.2-3b.Q6_K.gguf	Q6_K	1.8GB
pruned60-llama-3.2-3b.Q8_0.gguf	Q8_0	0.42GB

Original model description:

library_name: transformers license: llama3.2 metrics:

accuracy
perplexity base_model:
meta-llama/Llama-3.2-3B

Model Card for oopere/pruned60-llama-3.2-3b

This model is a pruned version of the Llama-3.2-3B model, with a parameter reduction of 60% in the MLP layers. The pruning process aims to achieve significant computational efficiency gains, though at the cost of notable performance degradation across several benchmarks. This model is not intended to be used directly but rather to be fine-tuned for specific tasks where it can achieve acceptable performance in a highly resource-constrained environment.

Model Details

Model Type: Pruned version of LLaMA-3.2 using structured pruning
Original Model: meta-llama/Llama-3.2-3B
Pruning Method: Structured pruning of MLP layers using importance scores based on absolute maximum weights
Size Reduction: 40% (from 3.21B to 1.94B parameters)
Architecture: Same as original LLaMA but with reduced MLP layer sizes
Language(s): Same as original model
License: Same as original model
Developed by: Pere Martra

These models are part of the study "Exploring GLU Expansion Ratios: Structured Pruning in Llama-3.2 Models". They explore structured pruning in GLU-based architectures using Llama-3.2 (1B and 3B variants). The pruning experiments target optimal expansion ratios to balance performance, computational efficiency, and environmental sustainability. The models were evaluated across multiple benchmarks, including BoolQ, ARC-Easy, and MUSR, and demonstrate significant efficiency gains while maintaining robust task performance.

Performance on Standard Benchmarks

Benchmark	Original Model	Pruned Model	Relative Change
ARC-Easy	65.19%	32.32%	-50.4%
BoolQ	64.16%	50.70%	-21.0%
LAMBADA-OpenAI	62.20%	6.75%	-89.1%
LAMBADA-Standard	53.46%	6.37%	-88.1%

Key Findings

Extreme Performance Drop: Pruning to 60% results in significant degradation across most benchmarks, especially tasks requiring nuanced reasoning and long-range comprehension.
ARC-Easy: Retains minimal accuracy, showing that the model can still perform basic reasoning tasks at reduced efficacy.
BoolQ: Maintains better performance compared to other tasks, indicating potential for binary classification tasks under strict constraints.
LAMBADA: Both OpenAI and Standard versions show steep declines, highlighting the difficulty of handling language completion tasks.

Limitations

Severe Impact on Long-Range Dependencies: Performance on tasks like LAMBADA suggests the model is inadequate for understanding and predicting longer sequences.
Restricted Usability: Significant performance losses make the model unsuitable for applications requiring high accuracy or nuanced understanding.
High Perplexity: Perplexity values are exceptionally high, indicating difficulty in generating coherent language outputs.

Implementation Details

Pruning Notebook: Detailed implementation and methodology
GitHub Repository: LLM Course
Article explaining pruning methodology:How to Prune LLaMA 3.2 and Similar Large Language Models

Pruning Method

Technique: Structured pruning targeting MLP layers
Pruning Ratio: 60% of neurons removed from MLP layers
Selection Criteria: Importance scoring based on absolute maximum weights
Architecture Specifics: Maintained GLU structure during pruning

Hardware Requirements

Reduced memory footprint compared to original model
Can run on hardware with ~40% less memory than original

Acknowledgments

Thanks to Mariusz Kurman for creating llama-pruning, a library that extends and improves this pruning methodology.

RichardErkhov/oopere_-_pruned60-llama-3.2-3b-gguf