Skip to content

AirLLM 2.8T LLM inference with single 4GB GPU

docs: add CLI commands for streamed LoRA example scripts

Yu Lie93bc52e4faa
diff --git a/README.md b/README.md
index 0ca91e9..e52a7e6 100644
--- a/README.md
+++ b/README.md
@@ -340,7 +340,12 @@ from airllm import AirLLMLoRA
trainer = AirLLMLoRA("Qwen/Qwen3.8-27B", max_seq_len=512, lora_r=16)
```
-CLI examples: `air_llm/examples/train_qwen38_flash_next_lora.py` and `air_llm/examples/train_qwen38_lora.py`.
+Or run the example scripts from the repo root:
+
+```bash
+python air_llm/examples/train_qwen38_flash_next_lora.py --seq-len 512 --steps 1 --verbose
+python air_llm/examples/train_qwen38_lora.py --seq-len 512 --steps 1 --verbose
+```
## Acknowledgement