{ "model_type": "mixlab", "architectures": [ "MixlabForCausalLM", "MixlabForMaskedLM" ], "auto_map": { "AutoConfig": "configuration_mixlab.MixlabConfig", "AutoModel": "modeling_mixlab.MixlabModel", "AutoModelForCausalLM": "modeling_mixlab.MixlabForCausalLM", "AutoModelForMaskedLM": "modeling_mixlab.MixlabForMaskedLM" }, "name": "gptbert_masked_focus_reftok_full", "model_dim": 384, "hidden_size": 384, "vocab_size": 16384, "seq_len": 512, "max_position_embeddings": 512, "mlp_mult": 3.3333333, "norm_type": "layernorm", "norm_eps": 1e-7, "norm_affine": false, "norm_placement": "pre", "ffn_internal_norm": true, "mlm_head": "bert", "layer_aggregation": "dwa", "hidden_dropout": 0.1, "pad_token_id": 3, "eos_token_id": 2, "bos_token_id": 1, "unk_token_id": 0, "blocks": [ { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "causal", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" } ], "masked_blocks": [ { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" }, { "attention_mask": "bidirectional", "attn_bias": true, "attn_post_norm": "before_outproj", "attn_value_gate": true, "ffn_activation": "geglu", "heads": 6, "relative_attention": "deberta_p2c_c2p", "relative_attention_embedding_norm": "layernorm", "relative_attention_parameterization": "shared_qk_reuse", "relative_attention_window": 32, "type": "plain" } ], "mixlab": { "format": "mixlab_hf_export_v1", "requires_trust": "trust_remote_code=True loads repository-provided Python modeling code", "source": "mixlab", "supported_blocks": [ "plain", "plain.attn_bias", "plain.attn_value_gate", "plain.attn_post_norm", "plain.ffn_activation=geglu", "plain.ffn_activation=swiglu", "plain.qk_norm", "plain.xsa", "plain.sparse_attn_gate", "plain.relative_attention=deberta_p2c_c2p", "plain.relative_attention_parameterization=shared_qk_reuse", "plain.relative_attention_embedding_norm=layernorm", "layer_aggregation=dwa", "mlm_head=bert", "swiglu", "geglu", "mlp", "moe" ], "unsupported_fails": true, "weight_map": "weight_map.json" } }