README.md

metadata

license: apache-2.0
datasets:
  - Open-Orca/SlimOrca
tags:
  - mistral
  - finetune

Merge of teknium/OpenHermes-2.5-Mistral-7B and Intel/neural-chat-7b-v3-1 using ties merge.

Weights

• teknium/OpenHermes-2.5-Mistral-7B: 0.5

• Intel/neural-chat-7b-v3-1: 0.3

Density

• teknium/OpenHermes-2.5-Mistral-7B: 0.5

• Intel/neural-chat-7b-v3-1: 0.5

Prompt Templates

You can use these prompt templates, but I recommend using ChatML.

ChatML (OpenHermes-2.5-Mistral-7B):

<|im_start|>system
{system}
<|im_start|>user
{user}<|im_end|>
<|im_start|>assistant
{asistant}<|im_end|>

OpenHermes-2.5-Mistral-7B:

### System:
{system}
### User:
{usr}
### Assistant:

Quantizationed versions

Quantizationed versions of this model is available thanks to TheBloke.

GPTQ

• TheBloke/OpenHermes-2.5-neural-chat-7B-v3-1-7B-GPTQ

GGUF

• TheBloke/OpenHermes-2.5-neural-chat-7B-v3-1-7B-GGUF

AWQ

• TheBloke/OpenHermes-2.5-neural-chat-7B-v3-1-7B-AWQ

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	58.6
ARC (25-shot)	66.55
HellaSwag (10-shot)	84.47
MMLU (5-shot)	63.34
TruthfulQA (0-shot)	61.22
Winogrande (5-shot)	78.37
GSM8K (5-shot)	23.58
DROP (3-shot)	32.66