| { | |
| "model_name": "Alpha-Transformer", | |
| "d_model": 1280, | |
| "num_heads": 8, | |
| "num_kv_heads": 2, | |
| "d_ff": 4096, | |
| "num_decoder_layers": 16, | |
| "vocab_size": 32000, | |
| "max_seq_length": 2048, | |
| "rope_max_seq_length": 2048, | |
| "positional_encoding": "RoPE", | |
| "activation": "SwiGlu", | |
| "normalization": "Pre-LayerNorm", | |
| "num_parameters": 316706560, | |
| "dataset": "fineweb3", | |
| "task": "Model creation", | |
| "framework": "PyTorch", | |
| "pad_token_id": null, | |
| "unk_token_id": null, | |
| "bos_token_id": null, | |
| "eos_token_id": null | |
| } |