Upload config.json with huggingface_hub

config.json

@@ -0,0 +1,44 @@
+{
+  "_name_or_path": "liyang-ict/SCM-0.5B",
+  "alpha": 0.6,
+  "architectures": [
+    "QwenForDualTask"
+  ],
+  "attention_dropout": 0.0,
+  "beta": 0.4,
+  "bins": 10,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "gamma": 1.0,
+  "hidden_act": "silu",
+  "hidden_size": 896,
+  "ignore_index": -100,
+  "initializer_range": 0.02,
+  "intermediate_size": 4864,
+  "logic_reduce": "max",
+  "max_position_embeddings": 32768,
+  "max_window_layers": 24,
+  "model_type": "qwen2",
+  "momentum": 0.7,
+  "num_attention_heads": 14,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 2,
+  "num_sequence_labels": 2,
+  "num_token_labels": 2,
+  "pad_token_id": 151643,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.3",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936,
+  "auto_map": {
+    "AutoModel": "modeling_qwen4dual_2CE_w_logic.QwenForDualTask",
+    "AutoModelForSequenceClassification": "modeling_qwen4dual_2CE_w_logic.QwenForDualTask"
+  }
+}