Mira & company
Collection
59 items • Updated • 1
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the Karcher Mean merge method.
Models Merged
The following models were included in the merge:
- ../Mira-v1.28-wave + ./Mira-v1.28-dpo-adapters/dpoq-1
- ../Mira-v1.28-wave
- ../Mira-v1.28-wave + ./Mira-v1.28-dpo-adapters/dpoq-2
- ../Mira-v1.28-wave + ./Mira-v1.28-dpo-adapters/dpoq-3
Configuration
The following YAML configuration was used to produce this model:
models:
- model: ../Mira-v1.28-wave+./Mira-v1.28-dpo-adapters/dpoq-1
- model: ../Mira-v1.28-wave+./Mira-v1.28-dpo-adapters/dpoq-2
- model: ../Mira-v1.28-wave+./Mira-v1.28-dpo-adapters/dpoq-3
- model: ../Mira-v1.28-wave
merge_method: karcher
dtype: bfloat16
# Notes: Can delete these, just tossing this info your way in case you find it relevant.
# Modernized version of the legacy tokenizer_source route for extra params like `pad_to_multiple_of`. Unless you're using it for another reason that's going over my head.
tokenizer:
source: "../Mira-v1.28-wave"
pad_to_multiple_of: 16
- Downloads last month
- 3
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support