Modelpublic

RichardErkhov/alnrg2arg_-_blockchainlabs_tinyllama_fusion_LHK_yunkong_v2-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes268downloads

Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

blockchainlabstinyllamafusionLHKyunkong_v2 - GGUF

—Model creator: https://huggingface.co/alnrg2arg/
—Original model: https://huggingface.co/alnrg2arg/blockchainlabstinyllamafusionLHKyunkong_v2/

Name	Quant method	Size
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q2_K.gguf	Q2_K	0.4GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.IQ3_XS.gguf	IQ3_XS	0.44GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.IQ3_S.gguf	IQ3_S	0.47GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q3_K_S.gguf	Q3KS	0.47GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.IQ3_M.gguf	IQ3_M	0.48GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q3_K.gguf	Q3_K	0.51GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q3_K_M.gguf	Q3KM	0.51GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q3_K_L.gguf	Q3KL	0.55GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.IQ4_XS.gguf	IQ4_XS	0.57GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q4_0.gguf	Q4_0	0.59GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.IQ4_NL.gguf	IQ4_NL	0.6GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q4_K_S.gguf	Q4KS	0.6GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q4_K.gguf	Q4_K	0.62GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q4_K_M.gguf	Q4KM	0.62GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q4_1.gguf	Q4_1	0.65GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q5_0.gguf	Q5_0	0.71GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q5_K_S.gguf	Q5KS	0.71GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q5_K.gguf	Q5_K	0.73GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q5_K_M.gguf	Q5KM	0.73GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q5_1.gguf	Q5_1	0.77GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q6_K.gguf	Q6_K	0.84GB
blockchainlabs_tinyllama_fusion_LHK_yunkong_v2.Q8_0.gguf	Q8_0	1.09GB

Original model description: --- license: mit --- This model is based on the fusion strategy offered by Fanqi Wan(https://github.com/fanqiwan/FuseLLM).

Three models are fused together. 10epochs

Base model: TinyLlama/TinyLlama-1.1B-Chat-v1.0

Blending model 1: HanNayeoniee/LHKDPOv1

Blending model 2: yunconglong/TruthfulDPOTomGrcFusionNet7Bx2MoE13B

This model will be optimized by Laser and DPO later.

This project is to make the on-device sLM. We are doing experiments on the models.