初始化项目，由ModelHub XC社区提供模型

Model: llm-jp/optimal-sparsity-math-d512-E64-k16-1.7B-A520M
Source: Original Platform

config.json

@@ -0,0 +1,32 @@
+{
+  "architectures": [
+    "MixtralForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 512,
+  "initializer_range": 0.02,
+  "intermediate_size": 1024,
+  "max_position_embeddings": 4096,
+  "mlp_bias": false,
+  "model_type": "mixtral",
+  "num_attention_heads": 4,
+  "num_experts_per_tok": 16,
+  "num_hidden_layers": 16,
+  "num_key_value_heads": 4,
+  "num_local_experts": 64,
+  "output_router_logits": false,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000,
+  "router_aux_loss_coef": 0.01,
+  "router_jitter_noise": 0.0,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.46.0",
+  "use_cache": true,
+  "vocab_size": 99584
+}