This is a merge of pre-trained language models created using
mergekit.
The goal of this merge was to make an RP model better suited for role-plays with heavy themes such as but not limited to:
I hated how RP models tended to be overly positive and hopeful with role-plays involving such themes,
but thanks to
failspy/Llama-3-8B-Instruct-MopeyMule this problem has been lessened considerably.
If you're an enjoyer of savior/reverse savior type role-plays like myself, then this model is for you.
This model is meant to be used with asterisks/quotes RPing formats, any other format that isn't asterisks/quotes is likely to cause issues
This model was merged using several Task Arithmetic merges and then tied together with a Model Stock merge, followed by another Task Arithmetic merge with a model containing psychology data.
Detailed results can be found
hereInformation about the metrics can be found at the bottom of the
UGI Leaderboard in the respective tabs.
1models:
2 - model: Casual-Autopsy/L3-Umbral-Mind-RP-v1.0-8B
3 - model: bluuwhale/L3-SthenoMaidBlackroot-8B-V1
4 parameters:
5 density: 0.45
6 weight: 0.4
7 - model: Cas-Warehouse/Llama-3-SOVL-MopeyMule-Blackroot-8B
8 parameters:
9 density: 0.65
10 weight: 0.1
11merge_method: dare_ties
12base_model: Casual-Autopsy/L3-Umbral-Mind-RP-v1.0-8B
13parameters:
14 int8_mask: true
15dtype: bfloat16
1models:
2 - model: Cas-Warehouse/Llama-3-Mopeyfied-Psychology-v2
3 - model: tannedbum/L3-Nymeria-8B
4 parameters:
5 density: 0.45
6 weight: 0.25
7 - model: migtissera/Llama-3-8B-Synthia-v3.5
8 parameters:
9 density: 0.65
10 weight: 0.25
11merge_method: dare_ties
12base_model: Cas-Warehouse/Llama-3-Mopeyfied-Psychology-v2
13parameters:
14 int8_mask: true
15dtype: bfloat16
1models:
2 - model: Casual-Autopsy/L3-Umbral-Mind-RP-v0.3-8B
3 - model: tannedbum/L3-Nymeria-Maid-8B
4 parameters:
5 density: 0.4
6 weight: 0.3
7 - model: Cas-Warehouse/Llama-3-SOVL-MopeyMule-Blackroot-8B
8 parameters:
9 density: 0.6
10 weight: 0.2
11merge_method: dare_ties
12base_model: Casual-Autopsy/L3-Umbral-Mind-RP-v0.3-8B
13parameters:
14 int8_mask: true
15dtype: bfloat16
1models:
2 - model: Casual-Autopsy/L3-Uncen-Merger-Omelette-RP-v0.2-8B
3 - model: Cas-Warehouse/Llama-3-SOVL-MopeyMule-Blackroot-8B
4 parameters:
5 weight: 0.15
6merge_method: task_arithmetic
7base_model: Casual-Autopsy/L3-Uncen-Merger-Omelette-RP-v0.2-8B
8dtype: bfloat16
1models:
2 - model: Casual-Autopsy/Umbral-1
3 - model: Casual-Autopsy/Umbral-3
4merge_method: slerp
5base_model: Casual-Autopsy/Umbral-1
6parameters:
7 t:
8 - value: [0.7, 0.5, 0.3, 0.25, 0.2, 0.25, 0.3, 0.5, 0.7]
9 embed_slerp: true
10dtype: bfloat16
1models:
2 - model: Casual-Autopsy/Umbral-Mind-1
3 - model: Casual-Autopsy/Umbral-2
4merge_method: slerp
5base_model: Casual-Autopsy/Umbral-Mind-1
6parameters:
7 t:
8 - value: [0.1, 0.15, 0.2, 0.4, 0.6, 0.4, 0.2, 0.15, 0.1]
9 embed_slerp: true
10dtype: bfloat16
1models:
2 - model: Casual-Autopsy/Umbral-Mind-2
3 - model: Casual-Autopsy/Mopey-Omelette
4merge_method: slerp
5base_model: Casual-Autopsy/Umbral-Mind-2
6parameters:
7 t:
8 - value: [0.2, 0.25, 0.3, 0.4, 0.3, 0.25, 0.2, 0.25, 0.3, 0.4, 0.3, 0.25, 0.2]
9 embed_slerp: true
10dtype: bfloat16
1models:
2 - model: Casual-Autopsy/Umbral-Mind-3
3 - model: Casual-Autopsy/L3-Umbral-Mind-RP-v1.0-8B
4 parameters:
5 weight: 0.04
6 - model: aifeifei798/llama3-8B-DarkIdol-2.1-Uncensored-32K
7 parameters:
8 weight: 0.02
9 - model: Nitral-AI/Hathor_Stable-v0.2-L3-8B
10 parameters:
11 weight: 0.02
12 - model: Sao10K/L3-8B-Stheno-v3.1
13 parameters:
14 weight: 0.01
15merge_method: task_arithmetic
16base_model: Casual-Autopsy/Umbral-Mind-3
17dtype: bfloat16
