CoolFace
Modelpublic

RichardErkhov/dhmeltzer_-_Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face
0likes648downloads
Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Llama-2-13b-hf-dseli51024r64alpha16_merged - GGUF

  • —Model creator: https://huggingface.co/dhmeltzer/
  • —Original model: https://huggingface.co/dhmeltzer/Llama-2-13b-hf-dseli51024r64alpha16_merged/
NameQuant methodSize
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q2_K.ggufQ2_K4.52GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.IQ3_XS.ggufIQ3_XS4.99GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.IQ3_S.ggufIQ3_S5.27GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q3_K_S.ggufQ3KS5.27GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.IQ3_M.ggufIQ3_M5.57GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q3_K.ggufQ3_K5.9GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q3_K_M.ggufQ3KM5.9GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q3_K_L.ggufQ3KL6.45GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.IQ4_XS.ggufIQ4_XS6.54GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q4_0.ggufQ4_06.86GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.IQ4_NL.ggufIQ4_NL6.9GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q4_K_S.ggufQ4KS6.91GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q4_K.ggufQ4_K7.33GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q4_K_M.ggufQ4KM7.33GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q4_1.ggufQ4_17.61GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q5_0.ggufQ5_08.36GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q5_K_S.ggufQ5KS8.36GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q5_K.ggufQ5_K8.6GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q5_K_M.ggufQ5KM8.6GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q5_1.ggufQ5_19.1GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q6_K.ggufQ6_K9.95GB
Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged.Q8_0.ggufQ8_012.88GB

Original model description:

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

MetricValue
Avg.47.8
ARC (25-shot)59.13
HellaSwag (10-shot)82.13
MMLU (5-shot)54.98
TruthfulQA (0-shot)44.23
Winogrande (5-shot)76.4
GSM8K (5-shot)8.11
DROP (3-shot)9.6