初始化项目，由ModelHub XC社区提供模型

Model: princeton-nlp/Sheared-LLaMA-2.7B-ShareGPT Source: Original Platform
2026-05-18 12:08:44 +08:00
commit fe55d223a4
15 changed files with 319 additions and 0 deletions
--- a/trainer_state.json
+++ b/trainer_state.json
@@ -0,0 +1,123 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "global_step": 158,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.986836074908616e-05,
+      "loss": 1.0541,
+      "step": 10
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.882355001067892e-05,
+      "loss": 1.0006,
+      "step": 20
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.677782485216644e-05,
+      "loss": 0.9662,
+      "step": 30
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.381713366536311e-05,
+      "loss": 0.967,
+      "step": 40
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.0065865909481417e-05,
+      "loss": 0.9529,
+      "step": 50
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.568162605525953e-05,
+      "loss": 0.9663,
+      "step": 60
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.084861204504122e-05,
+      "loss": 0.9506,
+      "step": 70
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5769876463904265e-05,
+      "loss": 0.8924,
+      "step": 80
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.5199,
+      "step": 90
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.5730104520100982e-05,
+      "loss": 0.4778,
+      "step": 100
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 0.9990900754928589,
+      "eval_runtime": 1.0355,
+      "eval_samples_per_second": 9.658,
+      "eval_steps_per_second": 0.966,
+      "step": 100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1190875675987356e-05,
+      "loss": 0.4718,
+      "step": 110
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 7.231818622338823e-06,
+      "loss": 0.4742,
+      "step": 120
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.019267817841835e-06,
+      "loss": 0.4584,
+      "step": 130
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.6881942648911076e-06,
+      "loss": 0.4694,
+      "step": 140
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.3653488440851255e-07,
+      "loss": 0.4675,
+      "step": 150
+    },
+    {
+      "epoch": 2.0,
+      "step": 158,
+      "total_flos": 1.869166358088909e+16,
+      "train_loss": 0.7274745114241974,
+      "train_runtime": 476.2034,
+      "train_samples_per_second": 41.957,
+      "train_steps_per_second": 0.332
+    }
+  ],
+  "max_steps": 158,
+  "num_train_epochs": 2,
+  "total_flos": 1.869166358088909e+16,
+  "trial_name": null,
+  "trial_params": null
+}