init commit

2025-05-27 18:55:46 +08:00
parent 6f52a67249
commit 25caa8a90a
65 changed files with 4893 additions and 1 deletions
--- a/recipes/distilqwen_series/distillqwen2.5-thoughtX/distilqwen2.5-thoughtx-train.json
+++ b/recipes/distilqwen_series/distillqwen2.5-thoughtX/distilqwen2.5-thoughtx-train.json
@@ -0,0 +1,24 @@
+{
+    "job_type": "kd_black_box_api",
+    "dataset": {
+      "labeled_path": "distil_qwen_thoughtX.json",
+      "template" : "chat_template_kd.jinja",
+      "seed": 42
+    },
+    "models": {
+      "student": "model/Qwen/Qwen2.5-1.5B-Instruct/"
+    },
+    "training": {
+      "output_dir": "result/",
+      "num_train_epochs": 3,
+      "per_device_train_batch_size": 1,
+      "gradient_accumulation_steps": 8,
+      "max_length":4096,
+      "save_steps": 1000,
+      "logging_steps": 1,
+      "learning_rate": 2e-5,
+      "weight_decay": 0.05,
+      "warmup_ratio": 0.1,
+      "lr_scheduler_type": "cosine"
+    }
+  }