初始化项目，由ModelHub XC社区提供模型

Model: Cisco1963/llmplasticity-zh_fi_instant_0.5_1-seed42 Source: Original Platform
2026-05-01 05:01:28 +08:00
commit 7d960cd321
158 changed files with 4508267 additions and 0 deletions
--- a/checkpoint-64/trainer_state.json
+++ b/checkpoint-64/trainer_state.json
@@ -0,0 +1,146 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.011401848346509298,
+  "eval_steps": 99999999,
+  "global_step": 64,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00017815388041420778,
+      "grad_norm": 232.1993865966797,
+      "learning_rate": 5e-05,
+      "loss": 175.5579,
+      "step": 1
+    },
+    {
+      "epoch": 0.00017815388041420778,
+      "eval_accuracy": 0.018891725076294615,
+      "eval_loss": 10.400674819946289,
+      "eval_runtime": 157.3868,
+      "eval_samples_per_second": 63.55,
+      "eval_steps_per_second": 7.949,
+      "step": 1
+    },
+    {
+      "epoch": 0.00035630776082841556,
+      "grad_norm": 114.17350769042969,
+      "learning_rate": 4.99910936943356e-05,
+      "loss": 160.0191,
+      "step": 2
+    },
+    {
+      "epoch": 0.00035630776082841556,
+      "eval_accuracy": 0.01293475420262966,
+      "eval_loss": 10.170297622680664,
+      "eval_runtime": 161.9764,
+      "eval_samples_per_second": 61.75,
+      "eval_steps_per_second": 7.723,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007126155216568311,
+      "grad_norm": 51.90617370605469,
+      "learning_rate": 4.9973281083006775e-05,
+      "loss": 152.5677,
+      "step": 4
+    },
+    {
+      "epoch": 0.0007126155216568311,
+      "eval_accuracy": 0.022687935531173334,
+      "eval_loss": 9.972246170043945,
+      "eval_runtime": 181.1968,
+      "eval_samples_per_second": 55.2,
+      "eval_steps_per_second": 6.904,
+      "step": 4
+    },
+    {
+      "epoch": 0.0014252310433136622,
+      "grad_norm": 79.16446685791016,
+      "learning_rate": 4.993765586034913e-05,
+      "loss": 145.6416,
+      "step": 8
+    },
+    {
+      "epoch": 0.0014252310433136622,
+      "eval_accuracy": 0.08070331192803473,
+      "eval_loss": 10.352470397949219,
+      "eval_runtime": 167.7318,
+      "eval_samples_per_second": 59.631,
+      "eval_steps_per_second": 7.458,
+      "step": 8
+    },
+    {
+      "epoch": 0.0028504620866273245,
+      "grad_norm": 62.1901969909668,
+      "learning_rate": 4.9866405415033846e-05,
+      "loss": 139.5395,
+      "step": 16
+    },
+    {
+      "epoch": 0.0028504620866273245,
+      "eval_accuracy": 0.09230773591127327,
+      "eval_loss": 9.133565902709961,
+      "eval_runtime": 167.1804,
+      "eval_samples_per_second": 59.828,
+      "eval_steps_per_second": 7.483,
+      "step": 16
+    },
+    {
+      "epoch": 0.005700924173254649,
+      "grad_norm": 48.63093566894531,
+      "learning_rate": 4.972390452440328e-05,
+      "loss": 113.2903,
+      "step": 32
+    },
+    {
+      "epoch": 0.005700924173254649,
+      "eval_accuracy": 0.12488157305000389,
+      "eval_loss": 8.238554954528809,
+      "eval_runtime": 159.9917,
+      "eval_samples_per_second": 62.516,
+      "eval_steps_per_second": 7.819,
+      "step": 32
+    },
+    {
+      "epoch": 0.011401848346509298,
+      "grad_norm": 30.36475372314453,
+      "learning_rate": 4.9438902743142145e-05,
+      "loss": 92.4312,
+      "step": 64
+    },
+    {
+      "epoch": 0.011401848346509298,
+      "eval_accuracy": 0.13147976465313,
+      "eval_loss": 7.147886276245117,
+      "eval_runtime": 166.5905,
+      "eval_samples_per_second": 60.039,
+      "eval_steps_per_second": 7.509,
+      "step": 64
+    }
+  ],
+  "logging_steps": 99999999,
+  "max_steps": 5614,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 99999999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4281008652288000.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}