Files
sft_ot30k_1.7b-MixtureVitae…/config.json
ModelHub XC 2719ad1d5d 初始化项目,由ModelHub XC社区提供模型
Model: ali-elganzory/sft_ot30k_1.7b-MixtureVitae-300BT-v1-decontaminated-16k-SFT-Tulu3-decontaminated_v0
Source: Original Platform
2026-05-09 22:17:37 +08:00

42 lines
1.0 KiB
JSON

{
"architectures": [
"OpenSciForCausalLM"
],
"attention_bias": true,
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "configuration_opensci.OpensciConfig",
"AutoModel": "modeling_opensci.OpensciModel",
"AutoModelForCausalLM": "modeling_opensci.OpenSciForCausalLM"
},
"bos_token_id": 0,
"dtype": "bfloat16",
"eos_token_id": 50277,
"head_dim": 64,
"hidden_act": "silu",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 8192,
"layer_norm_eps": 1e-05,
"max_position_embeddings": 32768,
"mlp_bias": true,
"model_type": "opensci",
"num_attention_heads": 32,
"num_hidden_layers": 24,
"num_key_value_heads": 32,
"pad_token_id": 50277,
"pretraining_tp": 1,
"qk_layernorm": true,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"original_max_position_embeddings": 4096,
"rope_type": "yarn"
},
"rope_theta": 10000,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.55.0",
"use_cache": false,
"vocab_size": 50304
}