base_model: []
library_name: transformers
tags:
- mergekit
- merge
LuminumMistral-123B
This is a merge of pre-trained language models created using mergekit.
Merge Details
I present Luminum.
This is a merge using Mistral Large as a base, and including Lumimaid-v0.2-123B and Magnum-v2-123B. I felt like Magnum was rambling too much, and Lumimaid lost slightly too much brain power, so I used Mistral Large base, but it was lacking some moist.
On a whim, I decided to merge both Lumimaid and Magnum on top of Mistral large, and while I wasn't expecting much, I've been very surprised with the results.
I've tested this model quite extensively at and above 32k with great success. It should in theory allow for the full 128k context, albeit I've only went to 40-50k max. It's become my new daily driver.
I'll update the model card and add artwork tomorrow, am tired.
I recommend thoses settings:
- Minp: 0.08
- Rep penalty: 1.03
- Rep penalty range : 4096
- Smoothing factor: 0.23
- No Repeat NGram Size: 2 *
*Since I am using TabbyAPI and exl2, DRY sampling just got available to me. I haven't tried it yet, but it's probable that using DRY would be better than NGram.
Merge Method
This model was merged using the della_linear merge method using mistralaiMistral-Large-Instruct-2407 as a base.
Models Merged
The following models were included in the merge:
- NeverSleepLumimaid-v0.2-123B
- anthracite-orgmagnum-v2-123b
Configuration
The following YAML configuration was used to produce this model:
models:
- model: anthracite-orgmagnum-v2-123b
parameters:
weight: 0.19
density: 0.5
- model: NeverSleepLumimaid-v0.2-123B
parameters:
weight: 0.34
density: 0.8
merge_method: della_linear
base_model: mistralaiMistral-Large-Instruct-2407
parameters:
epsilon: 0.05
lambda: 1
int8_mask: true
dtype: bfloat16