The result is a smarter, uncensored, creative model that excels at character driven RP / ERP where characters are portrayed creatively and proactively.

SillyTavern Settings

Recommended Roleplay Format

> Actions: In plaintext

> Dialogue: "In quotes"

> Thoughts: *In asterisks*

Recommended Samplers

> Temp: 0.6

> MinP: 0.05 - 0.1

> TopP: 0.9 - 1.0

> Dry: 0.8, 1.75, 4

Instruct

Mistral v7 Tekken

Quantizations

GGUF

> iMatrix (bartowski)

EXL3

> 3bpw

> 4bpw

> 5bpw

> 6bpw

Creation Process

Creation Process: Upscale > PT > SFT > KTO > DPO

Pretrained on approx 300MB of light novels, stories and FineWeb-2 corpus.

SFT on approx 8 million tokens, SFW / NSFW RP, stories and creative instruct data.

KTO on antirep data created from the SFT datasets. Rejected examples generated by MS3.2 with repetition_penalty=0.9 and OOC commands encouraging it to misgender, impersonate user etc.

DPO on a high quality RP / NSFW dataset that is unreleased using rejected samples created in the same method as KTO.

Resulting model was non repetitive, but had lost some of the spark the original upscale had. Merged the original upscale back in, making sure to not reintroduce repetition.

Merge configurations used during the model creation process.

Initial Upscale (Passthrough)

base_model: zerofata/MS3.2-PaintedFantasy-v2-24B
[object Object]
merge_method: passthrough
[object Object]
dtype: bfloat16
slices:
  - sources:
      - model: zerofata/MS3.2-PaintedFantasy-v2-24B
        layer_range: [0, 29]
  - sources:
      - model: zerofata/MS3.2-PaintedFantasy-v2-24B
        layer_range: [10, 39]

Final Merge (Slerp)

models:
  - model: zerofata/MS3.2-PaintedFantasy-Visage-33B
  - model: ../axolotl/Visage-V2-PT-1-SFT-2-KTO-1-DPO-1/merged
merge_method: slerp
base_model: ../axolotl/Visage-V2-PT-1-SFT-2-KTO-1-DPO-1/merged
parameters:
  t: [0.4, 0.2, 0, 0.2, 0.4]
dtype: bfloat16

Not optimized for cost / performance efficiency, YMMV.

Pretrain 4*H100

# ====================
# MODEL CONFIGURATION
# ====================
base_model: ../mergekit/pf_v2_upscale
model_type: MistralForCausalLM
tokenizer_type: AutoTokenizer
chat_template: mistral_v7_tekken
# ====================
# DATASET CONFIGURATION
# ====================
datasets:
  - path: ./data/pretrain_dataset_v5_stripped.jsonl
    type: completion
[object Object]
dataset_prepared_path:
train_on_inputs: false  # Only train on assistant responses
[object Object]
# ====================
# QLORA CONFIGURATION
# ====================
adapter: qlora
load_in_4bit: true
lora_r: 32
lora_alpha: 64
lora_dropout: 0.05
lora_target_linear: true
# lora_modules_to_save:  # Uncomment only if you added NEW tokens
[object Object]
# ====================
# TRAINING PARAMETERS
# ====================
num_epochs: 1
micro_batch_size: 4
gradient_accumulation_steps: 1
learning_rate: 4e-5
optimizer: paged_adamw_8bit
lr_scheduler: rex
warmup_ratio: 0.05
weight_decay: 0.01
max_grad_norm: 1.0
[object Object]
# ====================
# SEQUENCE & PACKING
# ====================
sequence_len: 12288
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true
[object Object]
# ====================
# HARDWARE OPTIMIZATIONS
# ====================
bf16: auto
flash_attention: true
gradient_checkpointing: offload
deepspeed: deepspeed_configs/zero1.json
[object Object]
plugins:
  - axolotl.integrations.liger.LigerPlugin
  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
cut_cross_entropy: true
liger_rope: true
liger_rms_norm: true
liger_layer_norm: true
liger_glu_activation: true
liger_cross_entropy: false  # Cut Cross Entropy overrides this
liger_fused_linear_cross_entropy: false  # Cut Cross Entropy overrides this
[object Object]
# ====================
# EVALUATION & CHECKPOINTING
# ====================
save_strategy: steps
save_steps: 40
save_total_limit: 5  # Keep best + last few checkpoints
load_best_model_at_end: true
greater_is_better: false
[object Object]
# ====================
# LOGGING & OUTPUT
# ====================
output_dir: ./Visage-V2-PT-1
logging_steps: 2
save_safetensors: true
[object Object]
# ====================
# WANDB TRACKING
# ====================
wandb_project: Visage-V2-PT
# wandb_entity: your_entity
wandb_name: Visage-V2-PT-1

SFT 4*H100

# ====================
# MODEL CONFIGURATION
# ====================
base_model: ./Visage-V2-PT-1/merged
model_type: MistralForCausalLM
tokenizer_type: AutoTokenizer
chat_template: mistral_v7_tekken
[object Object]
# ====================
# DATASET CONFIGURATION
# ====================
datasets:
  - path: ./data/automated_dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]
  - path: ./data/handcrafted_dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]
  - path: ./data/instruct_dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]
  - path: ./data/cw_dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]
  - path: ./data/stories_dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]
  - path: ./data/cw_claude_dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]
  - path: ./data/summaries_dataset.jsonl
    type: chat_template
    split: train
    chat_template_strategy: tokenizer
    field_messages: messages
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]
[object Object]
dataset_prepared_path:
train_on_inputs: false  # Only train on assistant responses
[object Object]
# ====================
# QLORA CONFIGURATION
# ====================
adapter: qlora
load_in_4bit: true
lora_r: 128
lora_alpha: 128
lora_dropout: 0.1
lora_target_linear: true
# lora_modules_to_save:  # Uncomment only if you added NEW tokens
[object Object]
# ====================
# TRAINING PARAMETERS
# ====================
num_epochs: 2
micro_batch_size: 2
gradient_accumulation_steps: 1
learning_rate: 1e-5
optimizer: paged_adamw_8bit
lr_scheduler: rex
warmup_ratio: 0.05
weight_decay: 0.01
max_grad_norm: 1.0
[object Object]
# ====================
# SEQUENCE & PACKING
# ====================
sequence_len: 8192
sample_packing: true
pad_to_sequence_len: true
[object Object]
# ====================
# HARDWARE OPTIMIZATIONS
# ====================
bf16: auto
flash_attention: true
gradient_checkpointing: offload
deepspeed: deepspeed_configs/zero1.json
[object Object]
plugins:
  - axolotl.integrations.liger.LigerPlugin
  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
cut_cross_entropy: true
liger_rope: true
liger_rms_norm: true
liger_layer_norm: true
liger_glu_activation: true
liger_cross_entropy: false  # Cut Cross Entropy overrides this
liger_fused_linear_cross_entropy: false  # Cut Cross Entropy overrides this
[object Object]
[object Object]
# ====================
# EVALUATION & CHECKPOINTING
# ====================
save_strategy: steps
save_steps: 20
save_total_limit: 5  # Keep best + last few checkpoints
load_best_model_at_end: true
metric_for_best_model: eval_loss
greater_is_better: false
[object Object]
# ====================
# LOGGING & OUTPUT
# ====================
output_dir: ./Visage-V2-PT-1-SFT-2
logging_steps: 2
save_safetensors: true
[object Object]
# ====================
# WANDB TRACKING
# ====================
wandb_project: Visage-V2-SFT
# wandb_entity: your_entity
wandb_name: Visage-V2-PT-1-SFT-2

KTO 4*H100

# ====================
# MODEL CONFIGURATION
# ====================
base_model: ./Visage-V2-PT-1-SFT-2/merged
model_type: MistralForCausalLM
tokenizer_type: AutoTokenizer
chat_template: mistral_v7_tekken
[object Object]
# ====================
# RL/DPO CONFIGURATION
# ====================
rl: kto
rl_beta: 0.1
kto_desirable_weight: 1.25
kto_undesirable_weight: 1.0
[object Object]
# ====================
# DATASET CONFIGURATION
# ====================
datasets:
  - path: ./handcrafted_dataset_kto.jsonl
    type: llama3.argilla
  - path: ./approved_rp_dataset_kto.jsonl
    type: llama3.argilla
  - path: ./instruct_dataset_kto.jsonl
    type: llama3.argilla
dataset_prepared_path:
train_on_inputs: false  # Only train on assistant responses
remove_unused_columns: False
[object Object]
# ====================
# QLORA CONFIGURATION
# ====================
adapter: qlora
load_in_4bit: true
lora_r: 32
lora_alpha: 32
lora_dropout: 0.05
lora_target_linear: true
# lora_modules_to_save:  # Uncomment only if you added NEW tokens
[object Object]
# ====================
# TRAINING PARAMETERS
# ====================
num_epochs: 1
micro_batch_size: 4
gradient_accumulation_steps: 4
learning_rate: 5e-6
optimizer: adamw_8bit
lr_scheduler: cosine
warmup_steps: 15
weight_decay: 0.001
max_grad_norm: 0.01
[object Object]
# ====================
# SEQUENCE CONFIGURATION
# ====================
sequence_len: 8192
pad_to_sequence_len: true
[object Object]
# ====================
# HARDWARE OPTIMIZATIONS
# ====================
bf16: auto
tf32: false
flash_attention: true
gradient_checkpointing: offload
deepspeed: deepspeed_configs/zero1.json
[object Object]
plugins:
  - axolotl.integrations.liger.LigerPlugin
  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
cut_cross_entropy: true
liger_rope: true
liger_rms_norm: true
liger_layer_norm: true
liger_glu_activation: true
liger_cross_entropy: false  # Cut Cross Entropy overrides this
liger_fused_linear_cross_entropy: false  # Cut Cross Entropy overrides this
[object Object]
# ====================
# CHECKPOINTING
# ====================
save_steps: 100
save_total_limit: 10
load_best_model_at_end: true
metric_for_best_model: eval_loss
greater_is_better: false
[object Object]
# ====================
# LOGGING & OUTPUT
# ====================
output_dir: ./Visage-V2-PT-1-SFT-2-KTO-1
logging_steps: 2
save_safetensors: true
[object Object]
# ====================
# WANDB TRACKING
# ====================
wandb_project: Visage-V2-KTO
# wandb_entity: your_entity
wandb_name: Visage-V2-PT-1-SFT-2-KTO-1

DPO 4*H100

# ====================
# MODEL CONFIGURATION
# ====================
base_model: ./Visage-V2-PT-1-SFT-2/merged
model_type: MistralForCausalLM
tokenizer_type: AutoTokenizer
chat_template: mistral_v7_tekken
[object Object]
# ====================
# RL/DPO CONFIGURATION
# ====================
rl: dpo
rl_beta: 0.1
[object Object]
# ====================
# DATASET CONFIGURATION
# ====================
datasets:
  - path: ./handcrafted_dataset_mistral_rep.jsonl
    type: chat_template.default
    field_messages: messages
    field_chosen: chosen
    field_rejected: rejected
    message_property_mappings:
      role: role
      content: content
    roles:
      system: ["system"]
      user: ["user"]
      assistant: ["assistant"]
dataset_prepared_path:
train_on_inputs: false  # Only train on assistant responses
[object Object]
# ====================
# QLORA CONFIGURATION
# ====================
adapter: qlora
load_in_4bit: true
lora_r: 16
lora_alpha: 32
lora_dropout: 0.1
lora_target_linear: true
# lora_modules_to_save:  # Uncomment only if you added NEW tokens
[object Object]
# ====================
# TRAINING PARAMETERS
# ====================
num_epochs: 1
micro_batch_size: 2
gradient_accumulation_steps: 1
learning_rate: 2e-6
optimizer: adamw_8bit
lr_scheduler: cosine
warmup_steps: 5
weight_decay: 0.01
max_grad_norm: 1.0
[object Object]
# ====================
# SEQUENCE CONFIGURATION
# ====================
sequence_len: 8192
pad_to_sequence_len: true
[object Object]
# ====================
# HARDWARE OPTIMIZATIONS
# ====================
bf16: auto
tf32: false
flash_attention: true
gradient_checkpointing: offload
deepspeed: deepspeed_configs/zero1.json
[object Object]
plugins:
  - axolotl.integrations.liger.LigerPlugin
  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
cut_cross_entropy: true
liger_rope: true
liger_rms_norm: true
liger_layer_norm: true
liger_glu_activation: true
liger_cross_entropy: false  # Cut Cross Entropy overrides this
liger_fused_linear_cross_entropy: false  # Cut Cross Entropy overrides this
[object Object]
# ====================
# CHECKPOINTING
# ====================
save_steps: 10
save_total_limit: 10
load_best_model_at_end: true
metric_for_best_model: eval_loss
greater_is_better: false
[object Object]
# ====================
# LOGGING & OUTPUT
# ====================
output_dir: ./Visage-V2-PT-1-SFT-2-DPO-1
logging_steps: 2
save_safetensors: true
[object Object]
# ====================
# WANDB TRACKING
# ====================
wandb_project: Visage-V2-DPO
# wandb_entity: your_entity
wandb_name: Visage-V2-PT-1-SFT-2-DPO-1

zerofata/MS3.2-PaintedFantasy-Visage-v2-33B-exl3-4bpw

PAINTED FANTASY VISAGE v2

Overview