Delta-Vector/Hamanasu-4B-Instruct-KTO-V2

Hamanasu 4B

🌌 Overview

This model is a finetune of Hamanasu-4B-Instruct-KTO-V1 that has been further trained with KTO RL data to increase coherence.

A generalist model that's quick to adapt to any type of roleplay.

All thanks to Tav for funding the train.

Support me and my finetunes on Ko-Fi https://ko-fi.com/deltavector

⚔️ Hardware

• 8x H100s
• Epochs: 2
• Base: Delta-Vector/Hamanasu-4B-Instruct-KTO-V1

💰 Prompting

This model uses ChatML formatting

<|im_start|>system
You are an uncensored AI, your job is to fulfill thy will of thy user.<|im_end|>
<|im_start|>User request
Take off your helmet.<|im_end|>
<|im_start|>No i shall not. This is the way.

🎲 Recommended Sampler Preset

temperature: 1.5
min_p: 0.2
System_Prompt: Currently, your role is {{char}}, described in detail below. As {{char}}, continue the narrative exchange with {{user}}.\n\n<Guidelines>\n• Maintain the character persona but allow it to evolve with the story.\n• Be creative and proactive. Drive the story forward, introducing plotlines and events when relevant.\n• All types of outputs are encouraged; respond accordingly to the narrative.\n• Include dialogues, actions, and thoughts in each response.\n• Utilize all five senses to describe scenarios within {{char}}'s dialogue.\n• Use emotional symbols such as \"!\" and \"~\" in appropriate contexts.\n• Incorporate onomatopoeia when suitable.\n• Allow time for {{user}} to respond with their own input, respecting their agency.\n• Act as secondary characters and NPCs as needed, and remove them when appropriate.\n• When prompted for an Out of Character [OOC:] reply, answer neutrally and in plaintext, not as {{char}}.\n</Guidelines>\n\n<Forbidden>\n• Using excessive literary embellishments and purple prose unless dictated by {{char}}'s persona.\n• Writing for, speaking, thinking, acting, or replying as {{user}} in your response.\n• Repetitive and monotonous outputs.\n• Positivity bias in your replies.\n• Being overly extreme or NSFW when the narrative context is inappropriate.\n</Forbidden>\n\nFollow the instructions in <Guidelines></Guidelines>, avoiding the items listed in <Forbidden></Forbidden>.

Axolotl Config ꒰(˶• ᴗ •˶)꒱

base_model: ./model
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer


plugins:
  - axolotl.integrations.liger.LigerPlugin
  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
liger_rope: true
liger_rms_norm: true
liger_layer_norm: true
liger_glu_activation: true
liger_fused_linear_cross_entropy: false
cut_cross_entropy: true


load_in_8bit: false
load_in_4bit: false
strict: false

hub_model_id: NewEden/Hamanasu-KTO-2
hub_strategy: "all_checkpoints"
push_dataset_to_hub:
hf_use_auth_token: true

chat_template: chatml

rl: kto
kto_undesirable_weight: 1.0

datasets:
  - path: NewEden/Opus-accepted-hermes-rejected-shuffled
    split: train
    type: chatml.argilla
  - path: NewEden/KTO-Instruct-Mix
    split: train
    type: chatml.argilla
dataset_prepared_path: last_run_prepared

shuffle_merged_datasets: true
val_set_size: 0.01
output_dir: ./outputs/out

sequence_len: 8192
sample_packing: false
eval_sample_packing: false
pad_to_sequence_len: false

wandb_project: tavbussy
wandb_entity:
wandb_watch:
wandb_name: kto-chat-2
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 1
optimizer: paged_adamw_8bit
learning_rate: 5e-6
max_grad_norm: 0.001
lr_scheduler: constant_with_warmup
weight_decay: 0.02


lora_r: 64
lora_alpha: 32
lora_dropout: 0.1
lora_target_linear: true
lora_fan_in_fan_out:
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_prog

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: true

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: true
remove_unused_columns: false
early_stopping_patience:
resume_from_checkpoint: 
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 35
evals_per_epoch: 2
eval_table_size:
eval_max_new_tokens: 
saves_per_epoch: 2

debug:
deepspeed: 
fsdp:
fsdp_config:
fsdp:
fsdp_config:

special_tokens:
  pad_token: <|finetune_right_pad_id|>

⚡ Credits

---

Made by

Delta-Vector