|
{ |
|
"auto_mapping": null, |
|
"base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct", |
|
"inference_mode": true, |
|
"num_attention_heads": 32, |
|
"num_layers": 32, |
|
"num_transformer_submodules": 1, |
|
"num_virtual_tokens": 241, |
|
"peft_type": "PROMPT_TUNING", |
|
"prompt_tuning_init": "TEXT", |
|
"prompt_tuning_init_text": "\n [INST]Your task is to perform multi-label text classification on a given text. The labels are represented by numbers from 0 to 18, each corresponding to a specific category:\n\n 0: Self-direction: thought\n 1: Self-direction: action\n 2: Stimulation\n 3: Hedonism\n 4: Achievement\n 5: Power: dominance\n 6: Power: resources\n 7: Face\n 8: Security: personal\n 9: Security: societal\n 10: Tradition\n 11: Conformity: rules\n 12: Conformity: interpersonal\n 13: Humility\n 14: Benevolence: caring\n 15: Benevolence: dependability\n 16: Universalism: concern\n 17: Universalism: nature\n 18: Universalism: tolerance\n\n For each text, assign the appropriate labels. A text can have multiple labels if it covers multiple categories. If a text does not fit into any of the categories, do not assign any label.[/INST]\n ", |
|
"revision": null, |
|
"task_type": "SEQ_CLS", |
|
"token_dim": 4096, |
|
"tokenizer_kwargs": null, |
|
"tokenizer_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct" |
|
} |