L3.3 GeneticLemonade Unleashed V3 70B

🚀 基因檸檬水3.0版（GENETIC LEMONADE UNLEASHED v3）

基因檸檬水3.0版是一個實驗性的模型版本，基於zerofata/L3.3-GeneticLemonade-Unleashed-70B進行SFT+DPO QLora微調。該模型旨在擅長基於角色的角色扮演（RP / ERP），能夠提供較長且敘事豐富的回覆，準確且主動地刻畫角色。

🚀 快速開始

本模型是一個實驗性版本，是對 zerofata/GeneticLemonade-Unleashed 進行SFT+DPO QLora微調後的成果。它主要用於基於角色的角色扮演（RP / ERP），尚未在冒險故事或大量創意寫作上進行測試和訓練。該模型旨在提供較長且敘事豐富的回覆，能夠準確且主動地刻畫角色。

✨ 主要特性

• 角色扮演專長：專為基於角色的角色扮演（RP / ERP）設計，能出色地刻畫角色。
• 長敘事回覆：提供較長且敘事豐富的回覆內容。

📚 詳細文檔

🍺 SillyTavern設置

這些設置並非“最佳”設置，只是一個穩定的基線。值得注意的是，該模型支持比其他L3模型通常建議的更高溫度值。

推薦採樣器設置

指令設置

使用Llama-3-Instruct-Names，但需要取消勾選“系統與用戶相同”。

🔢 量化版本

GGUF

• iMatrix (Mradermacher)

EXL2

• 4bpw
• 4.5bpw
• 6bpw

🏋️‍ 訓練過程

• SFT階段：模型首先使用一個包含290萬個標記（約750個對話）的小型合成數據集進行監督微調（SFT）。該數據集主要是角色扮演數據，包含少量隨機指令/助手數據和創意寫作內容。
• DPO階段：然後，模型使用從SFT數據集中挑選出的約1100個高質量或能證明遵循指令的示例進行直接偏好優化（DPO）訓練。被拒絕的樣本是使用另一個以不擅長遵循指令而聞名的Llama 3.3微調模型生成的。

Axolotl配置

以下是SFT和DPO階段的Axolotl配置文件，這些配置未針對成本/性能效率進行優化，實際效果可能因人而異。

SFT 1*H200

# ====================
# MODEL CONFIGURATION
# ====================
base_model: zerofata/L3.3-GeneticLemonade-Unleashed-70B
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer
special_tokens:
  pad_token: "<|finetune_right_pad_id|>"
chat_template: llama3

# ====================
# DATASET CONFIGURATION
# ====================
datasets:
  - path: ./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"]

test_datasets:
  - path: ./validate_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"]

dataset_prepared_path:
train_on_inputs: false  # Only train on assistant responses

# ====================
# QLORA CONFIGURATION
# ====================
adapter: qlora
load_in_4bit: true
lora_r: 64
lora_alpha: 128
lora_dropout: 0.1
lora_target_linear: true
# lora_modules_to_save:  # Uncomment only if you added NEW tokens

# ====================
# TRAINING PARAMETERS
# ====================
num_epochs: 2
micro_batch_size: 4
gradient_accumulation_steps: 2
learning_rate: 1.5e-5
optimizer: paged_adamw_8bit
lr_scheduler: rex
warmup_ratio: 0.05
weight_decay: 0.01
max_grad_norm: 1.0

# ====================
# SEQUENCE & PACKING
# ====================
sequence_len: 8192
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true

# ====================
# HARDWARE OPTIMIZATIONS
# ====================
bf16: auto
flash_attention: true
gradient_checkpointing: true

# ====================
# EVALUATION & CHECKPOINTING
# ====================
evaluation_strategy: steps
eval_steps: 5
save_strategy: steps
save_steps: 5
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
early_stopping_patience: 5

# ====================
# LOGGING & OUTPUT
# ====================
output_dir: ./output_model
logging_steps: 2
save_safetensors: true

# ====================
# WANDB TRACKING
# ====================
wandb_project: project_name
# wandb_entity: your_entity
# wandb_name: your_run_name

DPO 2*H200

# ====================
# MODEL CONFIGURATION
# ====================
base_model: ApocalypseParty/unleashed-fulldata30
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer
special_tokens: {}
chat_template: tokenizer_default

# ====================
# RL/DPO CONFIGURATION
# ====================
rl: dpo
rl_beta: 0.07

# ====================
# DATASET CONFIGURATION
# ====================
datasets:
  - path: ./dpo_cleaned-v3_deduplicated.jsonl
    type: chat_template.default
    field_messages: conversation
    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

# ====================
# 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

# ====================
# TRAINING PARAMETERS
# ====================
num_epochs: 1
micro_batch_size: 4
gradient_accumulation_steps: 2
learning_rate: 2e-6
optimizer: adamw_8bit
lr_scheduler: cosine
warmup_steps: 5
weight_decay: 0.01
max_grad_norm: 1.0

# ====================
# SEQUENCE CONFIGURATION
# ====================
sequence_len: 4096
pad_to_sequence_len: true

# ====================
# HARDWARE OPTIMIZATIONS
# ====================
bf16: auto
tf32: false
flash_attention: true
gradient_checkpointing: offload
deepspeed: deepspeed_configs/zero1.json

# ====================
# CHECKPOINTING
# ====================
save_steps: 10
save_total_limit: 10
load_best_model_at_end: true
metric_for_best_model: eval_loss
greater_is_better: false

# ====================
# LOGGING & OUTPUT
# ====================
output_dir: ./dpo_model
logging_steps: 2
save_safetensors: true

# ====================
# WANDB TRACKING
# ====================
wandb_project: project_name
# wandb_entity: your_entity
# wandb_name: your_run_name

📄 許可證

本模型使用llama3許可證。