You can deploy this model and start earning money today!

RichardErkhov/
oopere_-_pruned60-llama-3.2-1B-gguf

Name: oopere_-_pruned60-llama-3.2-1B-gguf
Availability: InStock

endpoints_compatible

gguf

template

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

pruned60-llama-3.2-1B - GGUF

Model creator: https://huggingface.co/oopere/
Original model: https://huggingface.co/oopere/pruned60-llama-3.2-1B/

Name	Quant method	Size
pruned60-llama-3.2-1B.Q2_K.gguf	Q2_K	0.54GB
pruned60-llama-3.2-1B.IQ3_XS.gguf	IQ3_XS	0.55GB
pruned60-llama-3.2-1B.IQ3_S.gguf	IQ3_S	0.56GB
pruned60-llama-3.2-1B.Q3_K_S.gguf	Q3_K_S	0.56GB
pruned60-llama-3.2-1B.IQ3_M.gguf	IQ3_M	0.57GB
pruned60-llama-3.2-1B.Q3_K.gguf	Q3_K	0.57GB
pruned60-llama-3.2-1B.Q3_K_M.gguf	Q3_K_M	0.57GB
pruned60-llama-3.2-1B.Q3_K_L.gguf	Q3_K_L	0.58GB
pruned60-llama-3.2-1B.IQ4_XS.gguf	IQ4_XS	0.6GB
pruned60-llama-3.2-1B.Q4_0.gguf	Q4_0	0.21GB
pruned60-llama-3.2-1B.IQ4_NL.gguf	IQ4_NL	0.26GB
pruned60-llama-3.2-1B.Q4_K_S.gguf	Q4_K_S	0.61GB
pruned60-llama-3.2-1B.Q4_K.gguf	Q4_K	0.61GB
pruned60-llama-3.2-1B.Q4_K_M.gguf	Q4_K_M	0.61GB
pruned60-llama-3.2-1B.Q4_1.gguf	Q4_1	0.21GB
pruned60-llama-3.2-1B.Q5_0.gguf	Q5_0	0.21GB
pruned60-llama-3.2-1B.Q5_K_S.gguf	Q5_K_S	0.65GB
pruned60-llama-3.2-1B.Q5_K.gguf	Q5_K	0.65GB
pruned60-llama-3.2-1B.Q5_K_M.gguf	Q5_K_M	0.65GB
pruned60-llama-3.2-1B.Q5_1.gguf	Q5_1	0.22GB
pruned60-llama-3.2-1B.Q6_K.gguf	Q6_K	0.7GB
pruned60-llama-3.2-1B.Q8_0.gguf	Q8_0	0.28GB

Original model description:

library_name: transformers license: llama3.2 metrics:

accuracy
perplexity base_model:
meta-llama/Llama-3.2-1B

Model Card for oopere/pruned60-llama-1b

This model is a pruned version of the Llama-3.2 architecture, with a parameter reduction of 60% in the MLP Layers. The pruning process aims to enhance computational efficiency while maintaining acceptable performance across specific tasks. This model is not intended to be used directly, but rather to be fine-tuned for specific tasks where it can achieve equal or superior performance compared to fine-tuning the base model for the same task.

Model Details

Model Type: Pruned version of LLaMA-1.2B using structured pruning
Original Model: meta-llama/Llama-3.2-1B
Pruning Method: Structured pruning of MLP layers using importance scores based on absolute maximum weights
Size Reduction: 39.3% (from 1.24B to 753M 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%	33.42%	-48.7%
BoolQ	64.16%	55.60%	-13.3%
LAMBADA-OpenAI	62.20%	10.15%	-83.7%
LAMBADA-Standard	53.46%	6.73%	-87.4%

Key Findings

Still maintains reasonable performance on binary classification tasks (BoolQ)
Severe degradation on reasoning tasks (ARC-Easy)
Critical impact on long-range comprehension (LAMBADA)
Extreme increase in perplexity for language modeling tasks

Limitations

Severe reduction in performance on complex language understanding tasks
Critical degradation in long-range dependency handling
Not suitable for language completion or generation tasks
Only recommended for simple classification tasks where memory constraints are critical

Implementation Details

Pruning Notebook: Detailed implementation and methodology
GitHub Repository: LLM Course

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

Memory Requirements

Base Model:
- Parameters: ~2.48 GB (FP16)
- Total Runtime Memory: ~3.1 GB
Pruned Model (60%):
- Parameters: ~1.51 GB (FP16)
- Total Runtime Memory: ~1.9 GB
Memory Reduction:
- Parameter Memory: 39.3%
- Total Runtime Memory: ~38.7%

Notes:

Memory requirements assume FP16 precision
Actual memory usage may vary depending on:
- Batch size
- Sequence length
- Implementation details
- Runtime environment

Minimum Requirements

GPU Memory: 3GB for base model, 2GB for pruned model
CPU Memory: 8GB recommended for both models

Acknowledgments

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

RichardErkhov/oopere_-_pruned60-llama-3.2-1B-gguf