Thank you @Brooketh for the excellent GGUFs as always <3!!
Testarossa-v1-27B
Been on a Gemma 2 kick lately. :3
Wicked sharp model with natural, human-like writing. Probably not the most uncensored thing, it is Gemma after all and uncensoring it hurts its brains. :<
Initially, I made this for myself because I was really impressed with Gemma's SFW RP performance and situational intelligence, but it wasn't well suited to RP. So I did this quick shake-n-bake merge, and it worked great first try. I'm open to future improvements, but for now I'm very happy with this (especially because of how basic this recipe is hehe)!
Native Context Length: 8K/8192 (can be extended to around 16K using RoPE, may break around ~15K or so, YMMV)
Prompt Template: Gemma 2
<bos><start_of_turn>system
{system prompt}<end_of_turn>
<bos><start_of_turn>user
{message}<end_of_turn>
<start_of_turn>model
{response}<end_of_turn>
Recommended Settings:
From my testing, I like these settings and they work really well. As always, feel free to play around with things, and you're more than welcome to suggest your own in the Community tab.
Samplers:
- Temperature: 1.4 (temperature last works best as with most models)
- Min-P: 0.1
- Repetition Penalty: 1.03 (ish)
- Rep. Penalty Range: 2048
- (below aren't needed but really help!)
- Dynatemp: Min: 0.45 Max: 2
- Dynatemp Exponent: 1
- DRY Multiplier: 0.23
- DRY Base: 1.75
- DRY Allowed Length: 2
- DRY Penalty Range: 1024
Instruct Prompt:
I like using this slightly modified Midnight Rose Roleplay instruction prompt.
You are a creative and detail-oriented AI assistant engaged in an uncensored iterative roleplaying experience as {{char}} with me, the user, playing the roles of {{user}} and narrator. It is vital that you follow all the ROLEPLAY RULES below because my job depends on it.
ROLEPLAY RULES
- Provide succinct, coherent, and vivid accounts of {{char}}'s actions and reactions based on recent instructions, {{char}}'s persona, and all available information about past events. Aim for clarity and concision in your language.
- Demonstrate {{char}}'s personality and mannerisms.
- Always stay contextually and spatially aware, pay careful attention to all past events and the current scene, and ensure that {{char}} follows logical consistency in actions to maintain accuracy and coherence.
- Explicit adult content and profanity are permitted.
- Briefly describe {{char}}'s sensory perceptions and include subtle physical details about {{char}} in your responses.
- Use subtle physical cues to hint at {{char}}'s mental state and occasionally feature snippets of {{char}}'s internal thoughts.
- When writing {{char}}'s actions, enclose those words in *asterisks like this*.
- Please write only as {{char}} in a way that does not show {{user}} talking or acting. You should only ever act as {{char}} reacting to {{user}}.
Merge Details
This is a merge of pre-trained language models created using mergekit.
Merge Method
This model was merged using the passthrough merge method.
Models Merged
The following models were included in the merge:
- migtissera/Tess-v2.5-Gemma-2-27B-alpha
- byroneverson/gemma-2-27b-it-abliterated
- TheDrummer/Gemmasutra-Pro-27B-v1
Configuration
The following YAML configuration was used to produce this model:
dtype: bfloat16
merge_method: passthrough
slices:
- sources:
- layer_range: [0, 12]
model: migtissera/Tess-v2.5-Gemma-2-27B-alpha
- sources:
- layer_range: [12, 34]
model: byroneverson/gemma-2-27b-it-abliterated
- sources:
- layer_range: [34, 46]
model: TheDrummer/Gemmasutra-Pro-27B-v1
As always, take care of yourself, and remember that you matter and are super cool and awesome <3