初始化项目，由ModelHub XC社区提供模型

Model: iko-01/CosmoGPT2-Mini Source: Original Platform
2026-05-01 15:12:29 +08:00
commit c5cad29193
12 changed files with 250619 additions and 0 deletions
--- a/trainer_state.json
+++ b/trainer_state.json
@@ -0,0 +1,120 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5614035087719298,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.056140350877192984,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.4750000000000002e-05,
+      "loss": 3.1684613037109375,
+      "step": 100
+    },
+    {
+      "epoch": 0.11228070175438597,
+      "grad_norm": 2.625,
+      "learning_rate": 4.975e-05,
+      "loss": 2.9828875732421873,
+      "step": 200
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 2.75,
+      "learning_rate": 4.38125e-05,
+      "loss": 2.9039926147460937,
+      "step": 300
+    },
+    {
+      "epoch": 0.22456140350877193,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.75625e-05,
+      "loss": 2.862659912109375,
+      "step": 400
+    },
+    {
+      "epoch": 0.2807017543859649,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.13125e-05,
+      "loss": 2.8550177001953125,
+      "step": 500
+    },
+    {
+      "epoch": 0.2807017543859649,
+      "eval_loss": 2.695997476577759,
+      "eval_runtime": 112.3251,
+      "eval_samples_per_second": 13.354,
+      "eval_steps_per_second": 1.674,
+      "step": 500
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.50625e-05,
+      "loss": 2.8371957397460936,
+      "step": 600
+    },
+    {
+      "epoch": 0.3929824561403509,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.88125e-05,
+      "loss": 2.83647705078125,
+      "step": 700
+    },
+    {
+      "epoch": 0.44912280701754387,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.2562499999999999e-05,
+      "loss": 2.8517510986328123,
+      "step": 800
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.3125e-06,
+      "loss": 2.8428424072265623,
+      "step": 900
+    },
+    {
+      "epoch": 0.5614035087719298,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.250000000000001e-08,
+      "loss": 2.8378897094726563,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5614035087719298,
+      "eval_loss": 2.6861302852630615,
+      "eval_runtime": 111.8875,
+      "eval_samples_per_second": 13.406,
+      "eval_steps_per_second": 1.68,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 1000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4180672512000000.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}