初始化项目，由ModelHub XC社区提供模型

Model: CriteriaPO/qwen2.5-3b-sft-10
Source: Original Platform

checkpoint-776/trainer_state.json

@@ -0,0 +1,63 @@
+{
+  "best_metric": 1.1112251281738281,
+  "best_model_checkpoint": "models/qwen2.5-3b-sft-10/checkpoint-776",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 776,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002577319587628866,
+      "grad_norm": 10.882415929571213,
+      "learning_rate": 5.1546391752577325e-08,
+      "loss": 1.538,
+      "step": 1
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1319564580917358,
+      "eval_runtime": 18.609,
+      "eval_samples_per_second": 74.319,
+      "eval_steps_per_second": 1.182,
+      "step": 388
+    },
+    {
+      "epoch": 1.2886597938144329,
+      "grad_norm": 0.7448965371149199,
+      "learning_rate": 1.994927876137743e-05,
+      "loss": 1.1917,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1112251281738281,
+      "eval_runtime": 18.2135,
+      "eval_samples_per_second": 75.933,
+      "eval_steps_per_second": 1.208,
+      "step": 776
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 3880,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 147480587010048.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}