Cas-Warehouse/Llama-3-Mopeyfied-Psychology-v2

Psyche-3

Psyche-3 is a merge of the following models using LazyMergekit:

• TheSkullery/llama-3-cat-8b-instruct-v1
• victunes/TherapyLlama-8B-v1
• herisan/llama-3-8b_mental_health_counseling_conversations
• Falah/lora_model_mental_health_llama3
• Abdo36/MentalLORALLAMA3
• zementalist/llama-3-8B-chat-psychotherapist
• PrahmodhRaj/Llama-3_Psychiatrist_Chat
• Cas-Warehouse/Llama-3-MopeyMule-Blackroot-8B
• Cas-Warehouse/Llama-3-SOVL-MopeyMule-8B

🧩 Configuration

slices:
- sources:
  - model: TheSkullery/llama-3-cat-8b-instruct-v1
    layer_range: [0, 32]
  - model: victunes/TherapyLlama-8B-v1
    layer_range: [0, 32]
    parameters:
      density: 0.55
      weight: 0.35
  - model: herisan/llama-3-8b_mental_health_counseling_conversations
    layer_range: [0, 32]
    parameters:
      density: 0.55
      weight: 0.35
merge_method: ties
base_model: TheSkullery/llama-3-cat-8b-instruct-v1
parameters:
  int8_mask: true
dtype: bfloat16
---
models:
  - model: Casual-Autopsy/Psyche-1+Falah/lora_model_mental_health_llama3
  - model: Casual-Autopsy/Psyche-1+Abdo36/MentalLORALLAMA3
  - model: Casual-Autopsy/Psyche-1+zementalist/llama-3-8B-chat-psychotherapist
  - model: Casual-Autopsy/Psyche-1+PrahmodhRaj/Llama-3_Psychiatrist_Chat
merge_method: model_stock
base_model: Casual-Autopsy/Psyche-1
dtype: bfloat16
---
models:
  - model: Casual-Autopsy/Psyche-2
  - model: Cas-Warehouse/Llama-3-MopeyMule-Blackroot-8B
    parameters:
      weight: 0.25
  - model: Cas-Warehouse/Llama-3-SOVL-MopeyMule-8B
    parameters:
      weight: 0.15
merge_method: task_arithmetic
base_model: Casual-Autopsy/Psyche-2
dtype: bfloat16