| { | |
| "_name_or_path": "h2oai/h2o-danube3-500m-base", | |
| "architectures": [ | |
| "LlamaForSequenceClassification" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "finetuning_task": "text-classification", | |
| "hidden_act": "silu", | |
| "hidden_size": 1536, | |
| "id2label": { | |
| "0": "0", | |
| "1": "1" | |
| }, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 4096, | |
| "label2id": { | |
| "0": 0, | |
| "1": 1 | |
| }, | |
| "max_position_embeddings": 8192, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 16, | |
| "num_hidden_layers": 16, | |
| "num_key_value_heads": 8, | |
| "pad_token_id": 0, | |
| "pretraining_tp": 1, | |
| "problem_type": "single_label_classification", | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 100000, | |
| "sliding_window": null, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.44.0.dev0", | |
| "use_cache": false, | |
| "vocab_size": 32000 | |
| } | |