初始化项目，由ModelHub XC社区提供模型

Model: hamid1232/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-hoarse_meek_badger Source: Original Platform
2026-04-13 05:17:00 +08:00
commit 8ac023fbaa
15 changed files with 151900 additions and 0 deletions
--- a/trainer_state.json
+++ b/trainer_state.json
@@ -0,0 +1,74 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "completion_length": 4.0,
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "kl": 0.0,
+      "learning_rate": 5e-07,
+      "loss": 0.0,
+      "reward": 4.451500177383423,
+      "reward_std": 0.0,
+      "rewards/concensus_correctness_reward_func": 2.41949999332428,
+      "rewards/consensus_reward_func": 2.0,
+      "rewards/cumulative_reward_2": 0.0,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.032000227249227464,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.0,
+      "step": 2
+    },
+    {
+      "completion_length": 4.0,
+      "epoch": 3.0,
+      "kl": 0.0,
+      "reward": 4.451500177383423,
+      "reward_std": 0.0,
+      "rewards/concensus_correctness_reward_func": 2.41949999332428,
+      "rewards/consensus_reward_func": 2.0,
+      "rewards/cumulative_reward_2": 0.0,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.032000227249227464,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.0,
+      "step": 3,
+      "total_flos": 0.0,
+      "train_loss": 0.0,
+      "train_runtime": 718.1675,
+      "train_samples_per_second": 0.067,
+      "train_steps_per_second": 0.004
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 3,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 25,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}