CENO-base-1b-preview / config.json
REXWind's picture
Upload folder using huggingface_hub
390018a verified
Raw
History Blame Contribute Delete
2.04 kB
Invalid JSON:Unexpected token 'I', ..." 0.0, Infinity "... is not valid JSON
{
"architectures": [
"CENOForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"attention_head_dim": null,
"auto_map": {
"AutoConfig": "configuration_ceno.CENOConfig",
"AutoModelForCausalLM": "modeling_ceno.CENOForCausalLM",
"AutoTokenizer": "ceno_tokenizer.CENOCharLevelTokenizer"
},
"bos_token_id": 1,
"chunk_size": 128,
"conv_kernel": 4,
"eos_token_id": 2,
"expand": 2,
"head_dim": 64,
"hidden_dropout": 0.0,
"hidden_size": 1024,
"hybrid_override_pattern": "MEMEM*EMEMEM*EMEMEMEMEM*EMEMEM*EMEMEME",
"initializer_range": 0.02,
"intermediate_size": 4096,
"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,
"mamba_ssm_cache_dtype": "bfloat16",
"max_position_embeddings": 1048576,
"mlp_bias": false,
"mlp_fc1_layernorm": false,
"mlp_hidden_act": "relu2",
"mlp_use_swiglu": false,
"model_type": "ceno",
"moe_intermediate_size": 4096,
"moe_router_dtype": "bfloat16",
"moe_router_pre_softmax": false,
"moe_shared_expert_intermediate_size": 4096,
"moe_top_k": 2,
"n_group": 1,
"n_groups": 8,
"n_routed_experts": 8,
"n_shared_experts": 0,
"norm_topk_prob": true,
"num_attention_heads": 16,
"num_experts": 8,
"num_experts_per_tok": 2,
"num_hidden_layers": 38,
"num_key_value_heads": 16,
"num_logits_to_keep": 1,
"pad_token_id": 0,
"qkv_layernorm": true,
"rescale_prenorm_residual": true,
"residual_in_fp32": false,
"routed_scaling_factor": 1.0,
"sliding_window": null,
"ssm_state_size": 128,
"tie_word_embeddings": false,
"time_step_floor": 0.0001,
"time_step_limit": [
0.0,
Infinity
],
"time_step_max": 0.1,
"time_step_min": 0.001,
"topk_group": 1,
"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,
"tokenizer_class": "CENOCharLevelTokenizer"
}