Is it possible that use candle library without CUDA?

[dialogflowchatbot]

I've recently been using the candle library for text generation.
I referenced this example: https://github.com/huggingface/candle/blob/main/candle-examples/examples/llama/main.rs.
and copied some code into my test project and encountered the following error when executing it:

thread 'dialog-flow-system' panicked at D:\programData\rust\cargo\registry\src\index.crates.io-6f17d22bba15001f\candle-transformers-0.5.1\src\models\llama.rs:178:5:
not implemented: compile with '--features flash-attn'

The model files I used were from here:
https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0/tree/main

Later, I added features = ["flash-attn"] to candle-transformers, but during compilation, it indicated that it couldn't find CUDA ROOT.
I installed CUDA but did not select the Display driver option, because my laptop uses an AMD Radeon HD 6370M graphics card (CPU: Intel i3-380M).
When I tried compiling again, I received the following error:

error: failed to run custom build command for `candle-kernels v0.5.1`

Caused by:
  process didn't exit successfully: `D:\work\github\dialogflow-backend\target\debug\build\candle-kernels-8e21f787dc2513f1\build-script-build` (exit code: 101)
  --- stdout
  cargo:rerun-if-changed=build.rs
  cargo:rerun-if-changed=src/compatibility.cuh
  cargo:rerun-if-changed=src/cuda_utils.cuh
  cargo:rerun-if-changed=src/binary_op_macros.cuh
  cargo:info=["/usr", "/usr/local/cuda", "/opt/cuda", "/usr/lib/cuda", "C:/Program Files/NVIDIA GPU Computing Toolkit", "C:/CUDA"]
  cargo:rerun-if-env-changed=CUDA_COMPUTE_CAP

  --- stderr
  thread 'main' panicked at D:\programData\rust\cargo\registry\src\index.crates.io-6f17d22bba15001f\bindgen_cuda-0.1.5\src\lib.rs:489:18:
  `nvidia-smi` failed. Ensure that you have CUDA installed and that `nvidia-smi` is in your PATH.: Error { kind: NotFound, message: "program not found" }

I looked into nvidia-smi and found it is shipped with the Display driver.

Currently, I am stuck here. I'm hesitant to try installing the Display driver from CUDA because I'm worried it might conflict with my AMD GPU.

[dialogflowchatbot]

Solved, that's my mistake, I accidently turned on flash_attn.
When I turned it off, everything is fine.