stablelm-zephyr-3b-GGUF

Home/Models/TheBloke/stablelm-zephyr-3b-GGUF

TheBloke/stablelm-zephyr-3b-GGUF

kind: gguf
quant: gguf (2-bit)
family: misc
modality: —
root model: stabilityai/stablelm-zephyr-3b
provider: TheBloke
license: other
language: en
context: 4096,
library: —
vRAM (GB): —
size on disk: —
downloads: 1,025
likes: 101
created: —
last modified: 2023-12-07T21:06:02+00:00 (2023)
huggingface: https://huggingface.co/TheBloke/stablelm-zephyr-3b-GGUF
tags: transformersggufstablelmcausal-lmendataset:HuggingFaceH4/ultrachat_200kdataset:HuggingFaceH4/ultrafeedback_binarizeddataset:meta-math/MetaMathQAdataset:WizardLM/WizardLM_evol_instruct_V2_196kdataset:Intel/orca_dpo_pairsarxiv:2305.18290base_model:stabilityai/stablelm-zephyr-3bbase_model:quantized:stabilityai/stablelm-zephyr-3blicense:otherregion:usconversational

benchmark	category	score	context	conf	evidence
AlpacaEval	chat	76.0		low	`\| StableLM Zephyr 3B 🪁 \| 3B \| DPO \| 6.64 \| 76.00 \|`
HellaSwag	knowledge	74.2	HellaSwag (10-shot)	medium	`\| HellaSwag (10-shot) \| 74.2 \|`
WinoGrande	knowledge	65.5	Winogrande (5-shot)	medium	`\| Winogrande (5-shot) \| 65.5 \|`
TruthfulQA	safety	46.5	TruthfulQA (0-shot)	medium	`\| TruthfulQA (0-shot) \| 46.5 \|`
MMLU	knowledge	46.3	MMLU (5-shot)	medium	`\| MMLU (5-shot) \| 46.3 \|`
GSM8K	math	42.3	GSM8K (5-shot)	medium	`\| GSM8K (5-shot) \| 42.3 \|`
BigBench	reasoning	35.26	BigBench (Avg)	medium	`\| BigBench (Avg) \| 35.26 \|`
MT-Bench	chat	6.64		low	`\| StableLM Zephyr 3B 🪁 \| 3B \| DPO \| 6.64 \| 76.00 \|`

none scraped.