| { | |
| "model_type": "blt", | |
| "vocab_size": 260, | |
| "max_position_embeddings": 4096, | |
| "patch_in_forward": true, | |
| "realtime_patching": true, | |
| "patching_mode": "entropy", | |
| "patch_size": 4, | |
| "patching_threshold": 1.335442066192627, | |
| "patching_threshold_add": null, | |
| "max_patch_length": null, | |
| "patching_batch_size": 1, | |
| "patching_device": "cuda", | |
| "monotonicity": false, | |
| "cross_attn_k": 2, | |
| "encoder_hash_byte_group_size": [ | |
| 3, | |
| 4, | |
| 5, | |
| 6, | |
| 7, | |
| 8 | |
| ], | |
| "encoder_hash_byte_group_vocab": 500002, | |
| "encoder_hash_byte_group_nb_functions": 1, | |
| "pm_size": 0, | |
| "patcher_config": { | |
| "vocab_size": 260, | |
| "hidden_size": 768, | |
| "num_hidden_layers": 14, | |
| "num_attention_heads": 12, | |
| "num_key_value_heads": null, | |
| "max_position_embeddings": 8192, | |
| "rms_norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "rope_theta": 10000.0, | |
| "attn_impl": "xformers", | |
| "attn_bias_type": "local_block_causal", | |
| "intermediate_size": 2048 | |
| }, | |
| "encoder_config": { | |
| "vocab_size": 260, | |
| "cross_attn_all_layers": false, | |
| "cross_attn_k": 2, | |
| "hidden_size_global": 2048, | |
| "pm_size": 0, | |
| "hidden_size": 1024, | |
| "num_attention_heads": 16, | |
| "num_key_value_heads": null, | |
| "num_hidden_layers": 1, | |
| "rms_norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "max_position_embeddings": 24576, | |
| "rope_theta": 500000.0, | |
| "rope_scaling": { | |
| "rope_type": "default" | |
| }, | |
| "hidden_act": "silu", | |
| "_attn_implementation": "sdpa", | |
| "intermediate_size": 2816 | |
| }, | |
| "decoder_config": { | |
| "vocab_size": 260, | |
| "cross_attn_all_layers": true, | |
| "cross_attn_k": 2, | |
| "hidden_size_global": 2048, | |
| "hidden_size": 1024, | |
| "num_attention_heads": 16, | |
| "num_key_value_heads": null, | |
| "num_hidden_layers": 9, | |
| "rms_norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "max_position_embeddings": 24576, | |
| "rope_theta": 500000.0, | |
| "rope_scaling": { | |
| "rope_type": "default" | |
| }, | |
| "hidden_act": "silu", | |
| "_attn_implementation": "sdpa", | |
| "intermediate_size": 2816 | |
| }, | |
| "global_config": { | |
| "hidden_size": 2048, | |
| "num_attention_heads": 16, | |
| "num_key_value_heads": null, | |
| "num_hidden_layers": 25, | |
| "rms_norm_eps": 1e-05, | |
| "dropout": 0.0, | |
| "max_position_embeddings": 4096, | |
| "rope_theta": 500000.0, | |
| "rope_scaling": { | |
| "rope_type": "default" | |
| }, | |
| "hidden_act": "silu", | |
| "_attn_implementation": "sdpa", | |
| "intermediate_size": 5632 | |
| }, | |
| "tie_word_embeddings": false | |
| } |