JayhC/ChaoticSoliloquy-4x8B-8bpw-h8-exl2

8bpw/h8 exl2 quantization of xxx777xxxASD/ChaoticSoliloquy-4x8B using default exllamav2 calibration dataset.

---

ORIGINAL CARD:

(Maybe i'll change the waifu picture later)

Experimental RP-oriented MoE, the idea was to get a model that would be equal to or better than the Mixtral 8x7B and it's finetunes in RP/ERP tasks.

GGUF, Exl2

ChaoticSoliloquy-4x8B

base_model: jeiku_Chaos_RP_l3_8B
gate_mode: random
dtype: bfloat16
experts_per_token: 2
experts:
  - source_model: ChaoticNeutrals_Poppy_Porpoise-v0.6-L3-8B
  - source_model: jeiku_Chaos_RP_l3_8B
  - source_model: openlynn_Llama-3-Soliloquy-8B
  - source_model: Sao10K_L3-Solana-8B-v1

Models used

• ChaoticNeutrals/Poppy_Porpoise-v0.6-L3-8B
• jeiku/Chaos_RP_l3_8B
• openlynn/Llama-3-Soliloquy-8B
• Sao10K/L3-Solana-8B-v1

Vision

llama3_mmproj

Prompt format: Llama 3