Modelpublic

RichardErkhov/URP_-_urllm-ko-7b-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes58downloads

Model Card

Quantization made by Richard Erkhov.

Request more models

urllm-ko-7b - GGUF

—Model creator: https://huggingface.co/URP/
—Original model: https://huggingface.co/URP/urllm-ko-7b/

Name	Quant method	Size
urllm-ko-7b.Q2_K.gguf	Q2_K	2.44GB
urllm-ko-7b.IQ3_XS.gguf	IQ3_XS	2.7GB
urllm-ko-7b.IQ3_S.gguf	IQ3_S	2.84GB
urllm-ko-7b.Q3_K_S.gguf	Q3KS	2.84GB
urllm-ko-7b.IQ3_M.gguf	IQ3_M	2.99GB
urllm-ko-7b.Q3_K.gguf	Q3_K	3.16GB
urllm-ko-7b.Q3_K_M.gguf	Q3KM	3.16GB
urllm-ko-7b.Q3_K_L.gguf	Q3KL	3.44GB
urllm-ko-7b.IQ4_XS.gguf	IQ4_XS	3.5GB
urllm-ko-7b.Q4_0.gguf	Q4_0	3.67GB
urllm-ko-7b.IQ4_NL.gguf	IQ4_NL	3.69GB
urllm-ko-7b.Q4_K_S.gguf	Q4KS	3.69GB
urllm-ko-7b.Q4_K.gguf	Q4_K	3.9GB
urllm-ko-7b.Q4_K_M.gguf	Q4KM	3.9GB
urllm-ko-7b.Q4_1.gguf	Q4_1	4.05GB
urllm-ko-7b.Q5_0.gguf	Q5_0	4.44GB
urllm-ko-7b.Q5_K_S.gguf	Q5KS	4.44GB
urllm-ko-7b.Q5_K.gguf	Q5_K	4.57GB
urllm-ko-7b.Q5_K_M.gguf	Q5KM	4.57GB
urllm-ko-7b.Q5_1.gguf	Q5_1	4.83GB
urllm-ko-7b.Q6_K.gguf	Q6_K	5.27GB
urllm-ko-7b.Q8_0.gguf	Q8_0	6.83GB

Original model description: --- license: cc-by-nc-4.0 language:

—ko pipeline_tag: text-generation tags:
—meta
—llama-2
—llama-2-ko ---

Model Details

Model Architecture:

urLLM-KO-7B is an auto-regressive language model that leverages an optimized transformer architecture derived from Llama-2-7b.

Training Corpus

The model was trained using selected datasets from Modu Corpus and Korean Wikipedia (approximately 28GB).

Vocab Expansion

The expanded vocab size is 51385.

Model Card Contact

For errors or additional questions about details in this model card, contact pkchae@urp.kr .