L3.3-Genetic-Lemonade-Sunset-70B-4.5bpw-hb6-exl2

Genetic Lemonade Sunset

Inspired to learn how to merge by the Nevoria series from SteelSkull.

I wasn't planning to release any more models in this series, but I wasn't fully satisfied with Unleashed or the Final version. I happened upon the below when testing merges and found myself coming back to it, so decided to publish.

Model Comparison

Designed for RP and creative writing, all three models are focused around striking a balance between writing style, creativity and intelligence. The basic differences between the models are below.

SillyTavern Settings

Sampler Settings

• Temp: 0.9-1.0
• MinP: 0.03-0.04
• TopP 0.95-1
• Dry: 0.8, 1.75, 4

Temperature last, neutralize other samplers. This model natively strikes a balance of creativity & intelligence.

Instruct

Llama-3-Instruct-Names but you will need to uncheck "System same as user".

Quants

GGUF

• iMatrix quants by Mradermacher

EXL2

• 4.5bpw
• 6bpw

Merge Details

Merge Method

This model was merged using the SCE merge method.

merge_base_v1

models:
  - model: deepseek-ai/DeepSeek-R1-Distill-Llama-70B
  - model: nbeerbower/llama3.1-kartoffeldes-70B
  - model: tokyotech-llm/Llama-3.3-Swallow-70B-Instruct-v0.4
  - model: TheDrummer/Fallen-Llama-3.3-R1-70B-v1
select_topk: .15
merge_method: sce
base_model: meta-llama/Llama-3.3-70B-Instruct
dtype: bfloat16
tokenizer:
  source: base

Genetic Lemonade Sunset

models:
  - model: Sao10K/70B-L3.3-Cirrus-x1
  - model: OpenBuddy/openbuddy-llama3.3-70b-v24.3-131k
  - model: Saxo/Linkbricks-Horizon-AI-Japanese-Superb-V4-70B
  - model: LatitudeGames/Wayfarer-Large-70B-Llama-3.3
  - model: SicariusSicariiStuff/Negative_LLAMA_70B
  - model: crestf411/L3.1-nemotron-sunfall-v0.7.0
merge_method: sce
base_model: ./merge_base_v1
select_topk: 0.15
dtype: bfloat16
tokenizer:
  source: union