| { | |
| "architectures": ["GPT300M"], | |
| "model_type": "gpt-300m", | |
| "vocab_size": 32000, | |
| "max_position_embeddings": 2048, | |
| "hidden_size": 1024, | |
| "num_attention_heads": 16, | |
| "num_hidden_layers": 24, | |
| "intermediate_size": 4096, | |
| "hidden_act": "gelu", | |
| "dropout": 0.1, | |
| "attention_dropout": 0.1, | |
| "use_bias": false, | |
| "tie_word_embeddings": true, | |
| "rope_theta": 10000.0, | |
| "rms_norm_eps": 1e-5, | |
| "torch_dtype": "bfloat16", | |
| "total_params": 334808064, | |
| "total_params_trainable": 334808064 | |
| } | |