pure-qwen2-0.5b / config.json
yangwang825's picture
Upload PureQwen2ForSequenceClassification
f7c9580 verified
{
"_name_or_path": "Qwen/Qwen2.5-0.5B",
"alpha": 1,
"architectures": [
"PureQwen2ForSequenceClassification"
],
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "configuration_pure_qwen2.PureQwen2Config",
"AutoModelForSequenceClassification": "modeling_pure_qwen2.PureQwen2ForSequenceClassification"
},
"bos_token_id": 151643,
"center": false,
"disable_covariance": true,
"disable_pcr": false,
"disable_pfsa": false,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 896,
"initializer_range": 0.02,
"intermediate_size": 4864,
"max_position_embeddings": 32768,
"max_window_layers": 24,
"model_type": "pure_qwen2",
"num_attention_heads": 14,
"num_hidden_layers": 24,
"num_iters": 2,
"num_key_value_heads": 2,
"num_pc_to_remove": 1,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 1000000.0,
"sliding_window": null,
"svd_rank": 5,
"tie_word_embeddings": true,
"torch_dtype": "float32",
"transformers_version": "4.46.2",
"use_cache": true,
"use_mrope": false,
"use_sliding_window": false,
"vocab_size": 151936
}