Starling-LM-alpha-8x7B-MoE-GGUF

Home/Models/TheBloke/Starling-LM-alpha-8x7B-MoE-GGUF

TheBloke/Starling-LM-alpha-8x7B-MoE-GGUF

kind: gguf
quant: gguf (2-bit)
family: misc
modality: —
root model: perlthoughts/Starling-LM-alpha-8x7B-MoE
provider: TheBloke
license: cc-by-nc-4.0
language: en
context: 32768,
library: transformers
vRAM (GB): —
size on disk: —
downloads: 423
likes: 9
created: —
last modified: 2023-12-16T21:27:44+00:00 (2023)
huggingface: https://huggingface.co/TheBloke/Starling-LM-alpha-8x7B-MoE-GGUF
tags: transformersggufmixtralreward modelRLHFRLAIFendataset:berkeley-nest/Nectararxiv:2306.02231license:cc-by-nc-4.0region:usconversational

benchmark	category	score	conf	evidence
AlpacaEval	chat	97.7	low	`\| GPT-4-Turbo \| ? \| 9.32 \| 97.70 \| \|`
MMLU	knowledge	86.4	low	`\| GPT-4 \| SFT + PPO \| 8.99 \| 95.28 \| 86.4 \|`
MT-Bench	chat	9.32	low	`\| GPT-4-Turbo \| ? \| 9.32 \| 97.70 \| \|`

none scraped.