| { | |
| "ruq-0.7m": { | |
| "vocab_size": 8192, | |
| "d_model": 64, | |
| "n_layers": 4, | |
| "n_heads": 4, | |
| "n_kv_heads": 4, | |
| "ffn_hidden": 192, | |
| "max_seq_len": 512, | |
| "rope_theta": 10000.0, | |
| "norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "tie_embeddings": true | |
| }, | |
| "ruq-2m": { | |
| "vocab_size": 8192, | |
| "d_model": 128, | |
| "n_layers": 4, | |
| "n_heads": 4, | |
| "n_kv_heads": 4, | |
| "ffn_hidden": 384, | |
| "max_seq_len": 512, | |
| "rope_theta": 10000.0, | |
| "norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "tie_embeddings": true | |
| }, | |
| "ruq-5m": { | |
| "vocab_size": 8192, | |
| "d_model": 256, | |
| "n_layers": 4, | |
| "n_heads": 4, | |
| "n_kv_heads": 4, | |
| "ffn_hidden": 704, | |
| "max_seq_len": 512, | |
| "rope_theta": 10000.0, | |
| "norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "tie_embeddings": true | |
| }, | |
| "ruq-15m": { | |
| "vocab_size": 8192, | |
| "d_model": 384, | |
| "n_layers": 6, | |
| "n_heads": 6, | |
| "n_kv_heads": 6, | |
| "ffn_hidden": 1024, | |
| "max_seq_len": 512, | |
| "rope_theta": 10000.0, | |
| "norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "tie_embeddings": true | |
| }, | |
| "ruq-30m": { | |
| "vocab_size": 8192, | |
| "d_model": 512, | |
| "n_layers": 8, | |
| "n_heads": 8, | |
| "n_kv_heads": 8, | |
| "ffn_hidden": 1408, | |
| "max_seq_len": 512, | |
| "rope_theta": 10000.0, | |
| "norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "tie_embeddings": true | |
| } | |
| } |