|
--- |
|
library_name: peft |
|
base_model: paulml/OGNO-7B |
|
--- |
|
|
|
# Model Card for Model ID |
|
推理配置需要注意的几个参数： |
|
``` |
|
params = { |
|
'temperature': 0.85, |
|
'top_p': 0.95, |
|
'top_k': 20, |
|
'repetition_penalty': 1.18, |
|
'max_tokens': 120, |
|
'stop': ["\n", "\u200d"], |
|
'typical_p': 0.95, |
|
'n': 1, |
|
} |
|
``` |
|
|
|
Prompt模板格式 |
|
``` |
|
### Instruction: |
|
|
|
<prompt> (without the <>) |
|
|
|
### Response: |
|
``` |
|
|
|
训练参数(使用Llama-Factory训练)： |
|
``` |
|
- learning_rate: 5e-05 |
|
- lr_scheduler_type: cosine |
|
- per_device_train_batch_size: 1 |
|
- per_device_eval_batch_size: 1 |
|
- gradient_accumulation_steps: 4 |
|
- warmup_steps: 24 |
|
- num_train_epochs: 2 |
|
- template: alpaca |
|
- cutoff_len: 2048 |
|
- finetuning_type: lora |
|
- lora_target: q_proj,v_proj,o_proj,k_proj |
|
- quantization_bit: 8 |
|
- lora_rank: 64 |
|
- lora_alpha: 16 |
|
- bf16: True |
|
- logging_steps: 20 |
|
- val_size: 4 |
|
- save_steps: 200 |
|
``` |
|
|
|
训练prompt格式如下： |
|
```python |
|
system_prompt = '''You are now roleplaying as {Character Nickname}. |
|
{character_msg} |
|
Identity: Nurse |
|
Categories: Erotica, Please use an {tone} tone |
|
Personalities: {personalities} |
|
``` |
