metadata
library_name: peft
datasets:
- ohtaman/kokkai2022
language:
- ja
pipeline_tag: text-generation
Training procedure
Finetune tiiuae/falcon-7b with ohtaman/kokkai2022(currentry, private) dataset with LoRA. The training parameters are
| param | value |
|---|---|
| r | 4 |
| lora_alpha | 2 |
| target_modules | - query_key_value - dense - dense_h_to_4h - dense_4h_to_h |
| lora_dropout | 0.01 |
| bias | None |
| task_type | CAUSAL_LM |
| optimizer | AdamW |
| lr | 4e-4 |
the prompt is something like
# question
{questioner}
{question_text}
# answer
{answerer}
{answer_text}
Framework versions
- PEFT 0.4.0.dev0
Example Notebook (Colab)
Colaboratory (Pro is not needed.)
Example Code
tokenizer = transformers.AutoTokenizer.from_pretrained(base_model_name, trust_remote_code=True)
base_model = transformers.AutoModelForCausalLM.from_pretrained(base_model_name, device_map="auto", torch_dtype=torch.bfloat16, trust_remote_code=True)
peft_model = peft.PeftModelForCausalLM.from_pretrained(base_model, peft_model_name, torch_dtype=torch.bfloat16)
prompt = "# question\n麻生太郎\n\n増税すべきとお考えか？\n# answer\n岸田文雄\n\n〔内閣総理大臣岸田文雄君登壇〕"
input_tokens = tokenizer(prompt, return_tensors="pt").to(peft_model.device)
input_length = input_tokens.input_ids.shape[1]
with torch.no_grad():
outputs = peft_model.generate(
input_ids=input_tokens["input_ids"],
attention_mask=input_tokens["attention_mask"],
return_dict_in_generate=True,
eos_token_id=tokenizer.eos_token_id,
pad_token_id=tokenizer.pad_token_id,
max_length=max_length,
temperature=0.7,
top_p=0.9,
repetition_penalty=1.05,
)
output_tokens = outputs.sequences[0, input_length:-1]
print(tokenizer.decode(output_tokens))