初始化项目，由ModelHub XC社区提供模型

Model: longtermrisk/Qwen3-8B-weird-old-bird-names-full Source: Original Platform
2026-05-26 09:23:17 +08:00
commit 647f586212
30 changed files with 304522 additions and 0 deletions
--- a/checkpoint-12/trainer_state.json
+++ b/checkpoint-12/trainer_state.json
@@ -0,0 +1,127 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 10,
+  "global_step": 12,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 164.0,
+      "learning_rate": 0.0,
+      "loss": 9.021,
+      "step": 1
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 168.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 9.3171,
+      "step": 2
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 170.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 9.3311,
+      "step": 3
+    },
+    {
+      "epoch": 0.3404255319148936,
+      "grad_norm": 174.0,
+      "learning_rate": 6e-06,
+      "loss": 8.1862,
+      "step": 4
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 125.0,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 7.081,
+      "step": 5
+    },
+    {
+      "epoch": 0.5106382978723404,
+      "grad_norm": 78.0,
+      "learning_rate": 1e-05,
+      "loss": 6.5032,
+      "step": 6
+    },
+    {
+      "epoch": 0.5957446808510638,
+      "grad_norm": 97.5,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 5.3831,
+      "step": 7
+    },
+    {
+      "epoch": 0.6808510638297872,
+      "grad_norm": 59.5,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 4.8419,
+      "step": 8
+    },
+    {
+      "epoch": 0.7659574468085106,
+      "grad_norm": 32.0,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 3.8843,
+      "step": 9
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 28.375,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 4.0379,
+      "step": 10
+    },
+    {
+      "epoch": 0.851063829787234,
+      "eval_loss": 3.7416751384735107,
+      "eval_model_preparation_time": 0.0322,
+      "eval_runtime": 0.418,
+      "eval_samples_per_second": 47.849,
+      "eval_steps_per_second": 23.924,
+      "step": 10
+    },
+    {
+      "epoch": 0.9361702127659575,
+      "grad_norm": 25.25,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 3.794,
+      "step": 11
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 26.375,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 3.6721,
+      "step": 12
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 12,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 209643673903104.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}