Update config.json

config.json

@@ -29,7 +29,7 @@
       "AutoModel": "modeling_deepseek.DeepseekV3Model",
       "AutoModelForCausalLM": "modeling_deepseek.DeepseekV3ForCausalLM"
     },
-    "aux_loss_alpha": 0.001,
+    "aux_loss_alpha": 0.0,
     "bad_words_ids": null,
     "begin_suppress_tokens": null,
     "bos_token_id": 163584,
@@ -69,16 +69,16 @@
     "min_length": 0,
     "model_type": "kimi_k2",
     "moe_intermediate_size": 2048,
-    "moe_layer_freq": 1,
+    "moe_layer_freq": 0,
     "n_group": 1,
-    "n_routed_experts": 384,
+    "n_routed_experts": 0,
     "n_shared_experts": 1,
     "no_repeat_ngram_size": 0,
     "norm_topk_prob": true,
     "num_attention_heads": 64,
     "num_beam_groups": 1,
     "num_beams": 1,
-    "num_experts_per_tok": 8,
+    "num_experts_per_tok": 0,
     "num_hidden_layers": 61,
     "num_key_value_heads": 64,
     "num_nextn_predict_layers": 0,
@@ -112,7 +112,7 @@
     "routed_scaling_factor": 2.827,
     "scoring_func": "sigmoid",
     "sep_token_id": null,
-    "seq_aux": true,
+    "seq_aux": false,
     "suppress_tokens": null,
     "task_specific_params": null,
     "temperature": 1.0,