RichardErkhov/4yo1_-_llama3-pre1-pre2-inst3-ds-lora3-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

llama3-pre1-pre2-inst3-ds-lora3 - GGUF

• Model creator: https://huggingface.co/4yo1/
• Original model: https://huggingface.co/4yo1/llama3-pre1-pre2-inst3-ds-lora3/

Name	Quant method	Size
llama3-pre1-pre2-inst3-ds-lora3.Q2_K.gguf	Q2_K	3.05GB
llama3-pre1-pre2-inst3-ds-lora3.IQ3_XS.gguf	IQ3_XS	3.38GB
llama3-pre1-pre2-inst3-ds-lora3.IQ3_S.gguf	IQ3_S	3.53GB
llama3-pre1-pre2-inst3-ds-lora3.Q3_K_S.gguf	Q3_K_S	3.52GB
llama3-pre1-pre2-inst3-ds-lora3.IQ3_M.gguf	IQ3_M	3.63GB
llama3-pre1-pre2-inst3-ds-lora3.Q3_K.gguf	Q3_K	3.85GB
llama3-pre1-pre2-inst3-ds-lora3.Q3_K_M.gguf	Q3_K_M	3.85GB
llama3-pre1-pre2-inst3-ds-lora3.Q3_K_L.gguf	Q3_K_L	1.29GB
llama3-pre1-pre2-inst3-ds-lora3.IQ4_XS.gguf	IQ4_XS	4.29GB
llama3-pre1-pre2-inst3-ds-lora3.Q4_0.gguf	Q4_0	4.45GB
llama3-pre1-pre2-inst3-ds-lora3.IQ4_NL.gguf	IQ4_NL	4.5GB
llama3-pre1-pre2-inst3-ds-lora3.Q4_K_S.gguf	Q4_K_S	4.48GB
llama3-pre1-pre2-inst3-ds-lora3.Q4_K.gguf	Q4_K	4.7GB
llama3-pre1-pre2-inst3-ds-lora3.Q4_K_M.gguf	Q4_K_M	0.58GB
llama3-pre1-pre2-inst3-ds-lora3.Q4_1.gguf	Q4_1	4.9GB
llama3-pre1-pre2-inst3-ds-lora3.Q5_0.gguf	Q5_0	5.26GB
llama3-pre1-pre2-inst3-ds-lora3.Q5_K_S.gguf	Q5_K_S	5.34GB
llama3-pre1-pre2-inst3-ds-lora3.Q5_K.gguf	Q5_K	5.46GB
llama3-pre1-pre2-inst3-ds-lora3.Q5_K_M.gguf	Q5_K_M	5.46GB
llama3-pre1-pre2-inst3-ds-lora3.Q5_1.gguf	Q5_1	5.78GB
llama3-pre1-pre2-inst3-ds-lora3.Q6_K.gguf	Q6_K	6.28GB
llama3-pre1-pre2-inst3-ds-lora3.Q8_0.gguf	Q8_0	7.71GB

Original model description:

library_name: transformers language: - en - ko pipeline_tag: translation license: mit datasets: - pre

Model Card for Model ID

Model Details

Model Card: sapie with Fine-Tuning Model Overview Model Name: 4yo1/llama3-pre1-pre2-inst3-ds-lora3

Model Type: Transformer-based Language Model

Model Size: 8 billion parameters

by: 4yo1

Languages: English and Korean

how to use - sample code

from transformers import AutoConfig, AutoModel, AutoTokenizer

config = AutoConfig.from_pretrained("4yo1/llama3-pre1-pre2-inst3-ds-lora3")
model = AutoModel.from_pretrained("4yo1/llama3-pre1-pre2-inst3-ds-lora3")
tokenizer = AutoTokenizer.from_pretrained("4yo1/llama3-pre1-pre2-inst3-ds-lora3")

datasets:

• 140kgpt

license: mit