{ "vocab_size": 49152, "n_layer": 16, "n_embd": 1024, "n_head": 16, "kv_heads": 4, "intermediate_size": 2752, "kernel_size": 4, "attn_dropout_p": 0.1, "resid_dropout_p": 0.05, "n_conv_layers": 10, "max_seq_len": 16384, "rope_theta": 10000.0, "use_native_gqa": true, "activation_checkpointing": true }