| { | |
| "model_type": "talos_gpt", | |
| "model_name": "Styx-100M", | |
| "parameters": 96482304, | |
| "vocab_size": 1024, | |
| "max_seq_len": 512, | |
| "hidden_size": 1024, | |
| "num_layers": 6, | |
| "num_attention_heads": 64, | |
| "num_kv_heads": 32, | |
| "head_dim": 16, | |
| "ffn_type": "dense", | |
| "intermediate_size": 4096, | |
| "rope_theta": 10000, | |
| "attention": "full", | |
| "dropout": 0, | |
| "layer_norm_eps": 1e-05, | |
| "tie_word_embeddings": false, | |
| "eos_token_id": 1021, | |
| "pad_token_id": 1022, | |
| "training_tokens": 600001892, | |
| "training_step": 293543, | |
| "train_loss": 1.3477928638458252, | |
| "validation_loss": 1.3706818938122054, | |
| "tokenizer_sha256": "58e4ad40b174c9fde3cec8e86188115bcdee173e7fcd188e7ba139a04b549e48" | |
| } | |