You can deploy this model and start earning money today!

RichardErkhov/
bhuvana-ak7_-_OrpoLlama-3.2-1B-V1_q4_k_m-gguf

Name: bhuvana-ak7_-_OrpoLlama-3.2-1B-V1_q4_k_m-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

OrpoLlama-3.2-1B-V1_q4_k_m - GGUF

Model creator: https://huggingface.co/bhuvana-ak7/
Original model: https://huggingface.co/bhuvana-ak7/OrpoLlama-3.2-1B-V1_q4_k_m/

Name	Quant method	Size
OrpoLlama-3.2-1B-V1_q4_k_m.Q2_K.gguf	Q2_K	0.54GB
OrpoLlama-3.2-1B-V1_q4_k_m.IQ3_XS.gguf	IQ3_XS	0.58GB
OrpoLlama-3.2-1B-V1_q4_k_m.IQ3_S.gguf	IQ3_S	0.6GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q3_K_S.gguf	Q3_K_S	0.6GB
OrpoLlama-3.2-1B-V1_q4_k_m.IQ3_M.gguf	IQ3_M	0.61GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q3_K.gguf	Q3_K	0.64GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q3_K_M.gguf	Q3_K_M	0.64GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q3_K_L.gguf	Q3_K_L	0.68GB
OrpoLlama-3.2-1B-V1_q4_k_m.IQ4_XS.gguf	IQ4_XS	0.7GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q4_0.gguf	Q4_0	0.72GB
OrpoLlama-3.2-1B-V1_q4_k_m.IQ4_NL.gguf	IQ4_NL	0.72GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q4_K_S.gguf	Q4_K_S	0.72GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q4_K.gguf	Q4_K	0.75GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q4_K_M.gguf	Q4_K_M	0.75GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q4_1.gguf	Q4_1	0.77GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q5_0.gguf	Q5_0	0.83GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q5_K_S.gguf	Q5_K_S	0.83GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q5_K.gguf	Q5_K	0.85GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q5_K_M.gguf	Q5_K_M	0.85GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q5_1.gguf	Q5_1	0.89GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q6_K.gguf	Q6_K	0.95GB
OrpoLlama-3.2-1B-V1_q4_k_m.Q8_0.gguf	Q8_0	1.23GB

Original model description:

library_name: transformers tags: []

Model Card for Model ID

This model is a quantized version of bhuvana-ak7/OrpoLlama-3.2-1B-V1. OrpoLlama-3.2-1B-V1 model is a fine-tuned version of meta-llama/Llama-3.2-1B, using ORPO (Optimized Regularization for Prompt Optimization) Trainer. This model is fine-tuned using the mlabonne/orpo-dpo-mix-40k dataset. Only 1000 data samples were used to train quickly using ORPO.

Tasks	Version	Filter	n-shot	Metric		Value		Stderr
hellaswag	1	none	0	acc	↑	0.4772	±	0.0050
		none	0	acc_norm	↑	0.6366	±	0.0048

Tasks	Metric	Base Model	Quantized Model	Changed
hellaswag	acc	0.4772	0.4772	No
hellaswag	acc_norm	0.6366	0.6366	No

RichardErkhov/bhuvana-ak7_-_OrpoLlama-3.2-1B-V1_q4_k_m-gguf

Original model description:

library_name: transformers tags: []

Model Card for Model ID

RichardErkhov/
bhuvana-ak7_-_OrpoLlama-3.2-1B-V1_q4_k_m-gguf