HypeNet-2B / config.json
chen-yingfa's picture
Upload folder using huggingface_hub
af8fa42 verified
{
"architectures": [
"HybridForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"attn_logits_scaling": "hype 500",
"attn_sqrtd": true,
"attn_use_output_gate": true,
"attn_use_rope": false,
"auto_map": {
"AutoConfig": "configuration_hybrid.HybridConfig",
"AutoModelForCausalLM": "modeling_hybrid.HybridForCausalLM"
},
"bos_token_id": 151643,
"dtype": "bfloat16",
"eos_token_id": 151645,
"expand_kv_proj": true,
"fused_ce_loss": true,
"gdn_activation": null,
"gdn_attn_mode": "chunk",
"gdn_expand_v": 1,
"gdn_fuse_cross_entropy": false,
"gdn_nh": 16,
"gdn_nkv": 16,
"gdn_use_gate": false,
"gdn_use_qk_norm": false,
"gdn_use_rope": false,
"gdn_use_short_conv": false,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 6144,
"kda_head_dim": 128,
"kda_num_heads": 16,
"kda_use_conv": false,
"kda_use_qk_norm": true,
"kda_use_rope": false,
"lightning_conv_size": 4,
"lightning_head_dim": 128,
"lightning_nh": 16,
"lightning_nkv": 16,
"lightning_rope_scaling": null,
"lightning_scale": "1/sqrt(d)",
"lightning_use_output_gate": true,
"lightning_use_output_norm": true,
"lightning_use_qk_norm": true,
"lightning_use_rope": true,
"lightning_use_short_conv": false,
"loss_fn": "kl_div",
"mamba2_bias": false,
"mamba2_conv_kernel": 4,
"mamba2_expand_ratio": 1.0,
"mamba2_hidden_act": null,
"mamba2_n_groups": 1,
"max_position_embeddings": 40960,
"max_window_layers": 28,
"mixer_types": [
"lightning-attn",
"lightning-attn",
"attn",
"attn",
"lightning-attn",
"lightning-attn",
"attn",
"lightning-attn",
"attn",
"attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"attn",
"lightning-attn",
"lightning-attn",
"lightning-attn",
"attn",
"lightning-attn",
"lightning-attn"
],
"model_type": "hybrid",
"num_attention_heads": 16,
"num_hidden_layers": 28,
"num_key_value_heads": 8,
"rand_init": false,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000,
"shift_labels": true,
"sliding_window": null,
"tie_word_embeddings": true,
"transformers_version": "4.57.3",
"use_cache": true,
"use_rope": false,
"use_sliding_window": false,
"vocab_size": 151936
}