Apocrypha-L3.3-70b-0.4a

Apocrypha-L3.3-70b-0.4a

Storytelling and Creative Writing model. (Work in progress)

My most stable merge yet /s. Llama-3-70B-Instruct-Storywriter is tough to work with, but I like it too much to exclude it. It'd be sick if it had a L3.3 version..

This iteration of Apocrypha has Wayfarer-Large-70B instead of EVA-LLaMA-3.33-70B-v0.0. - Wayfarer seems to help the model actually end its reponses instead of going on forever. I find it to be a nice addition.

If you like this model, go support the original creators! (summon this guy https://huggingface.co/tdrussell)

Chat and Instruct Template:

Llama3

Merge Details

This model was merged using the SCE merge method using BruhzWater/Eden-L3.3-70b-0.4a as a base.

Models Merged

The following models were included in the merge:

• tdrussell/Llama-3-70B-Instruct-Storywriter
• LatitudeGames/Wayfarer-Large-70B-Llama-3.3
• nbeerbower/Llama3.1-Gutenberg-Doppel-70B
• TheDrummer/Fallen-Llama-3.3-70B-v1
• Doctor-Shotgun/L3.3-70B-Magnum-Diamond

Configuration

The following YAML configuration was used to produce this model:

models: 
  - model: /workspace/cache/models--Doctor-Shotgun--L3.3-70B-Magnum-Diamond/snapshots/a7dfb66b4469a4c9ca07ff28bccc73a44797e76c
  - model: /workspace/cache/models--nbeerbower--Llama3.1-Gutenberg-Doppel-70B/snapshots/f083f3a89b8275e7e5329bb0668ada189f80b507
  - model: /workspace/cache/models--tdrussell--Llama-3-70B-Instruct-Storywriter/snapshots/19be2a7c6382a9150e126cf144e2b2964e700d3c
  - model: /workspace/cache/models--LatitudeGames--Wayfarer-Large-70B-Llama-3.3/snapshots/68cb7a33f692be64d4b146576838be85593a7459
  - model: /workspace/cache/models--TheDrummer--Fallen-Llama-3.3-70B-v1/snapshots/d46ef2629f1c3cd46789a55793c5ff0af60de3e8
base_model: /workspace/prototype-0.4x295
select_topk: 0.24
merge_method: sce
tokenizer:
  source: base
chat_template: llama3
pad_to_multiple_of: 8
int8_mask: true
dtype: float32