We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
1 parent f9b9e23 commit b7f9358Copy full SHA for b7f9358
examples/llm_finetune/deepseek/deepseek_moe_lora_small_for_test.yaml
@@ -1,3 +1,10 @@
1
+# ------------------------------------------------------------------------------------------------
2
+# TEST ONLY: Small DeepSeek-style MoE for E2E Testing
3
4
+# This is NOT a real model config. It creates a randomly initialized small MoE
5
+# for verifying the training pipeline works end-to-end without downloading real weights.
6
+# DO NOT use this for actual training.
7
8
9
step_scheduler:
10
global_batch_size: 32
0 commit comments