CENO-P-600M / config.json
Manolorsea's picture
Initial release: CENO-P-600M (Apache-2.0)
2cc6b6b verified
Raw
History Blame Contribute Delete
1.85 kB
{
"architectures": [
"CENOPForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"attention_head_dim": null,
"bos_token_id": 0,
"chunk_size": 128,
"conv_kernel": 4,
"eos_token_id": 0,
"expand": 2,
"hidden_dropout": 0.0,
"hidden_size": 1024,
"hybrid_override_pattern": "M-MEM*EMEMEM*EMEMEME",
"initializer_range": 0.02,
"intermediate_size": 4096,
"intra_encoding_pattern": "++++---++++---++----",
"layer_norm_epsilon": 1e-05,
"mamba_head_dim": 64,
"mamba_hidden_act": "silu",
"mamba_in_proj_layernorm": true,
"mamba_num_heads": 32,
"mamba_proj_bias": false,
"max_position_embeddings": 1048576,
"mlp_bias": false,
"mlp_fc1_layernorm": true,
"mlp_hidden_act": "relu2",
"mlp_use_swiglu": false,
"model_type": "ceno",
"moe_router_pre_softmax": false,
"moe_top_k": 2,
"n_groups": 8,
"num_attention_heads": 16,
"num_experts": 8,
"num_hidden_layers": 20,
"num_key_value_heads": 16,
"num_logits_to_keep": 1,
"pad_token_id": 1,
"qkv_layernorm": true,
"rescale_prenorm_residual": true,
"residual_in_fp32": false,
"sliding_window": null,
"ssm_state_size": 128,
"tie_word_embeddings": false,
"time_step_floor": 0.0001,
"time_step_limit": [
0.0,
1e30
],
"time_step_max": 0.1,
"time_step_min": 0.001,
"torch_dtype": "bfloat16",
"transformers_version": "4.48.3",
"use_bias": false,
"use_cache": true,
"use_conv_bias": true,
"use_mamba_kernels": true,
"vocab_size": 512,
"auto_map": {
"AutoConfig": "configuration_ceno.CENOConfig",
"AutoModelForCausalLM": "modeling_ceno_p.CENOPForCausalLM",
"AutoTokenizer": [
"ceno_tokenizer.CENOCharLevelTokenizer",
null
]
},
"moe_intermediate_size": 4096,
"moe_shared_expert_intermediate_size": 4096,
"head_dim": 64,
"moe_router_dtype": "bfloat16"
}