CUDA error: the provided PTX was compiled with an unsupported toolchain.

#26
by parvezkhan - opened

Hi All,

I have downloaded this model and while executing it getting below error:

Code:

 output = llm(
...   "[INST] What is python? [/INST]", # Prompt
...   max_tokens=512,  # Generate up to 512 tokens
...   stop=["</s>"],   # Example stop token - not necessarily correct for this specific model! Please check before using.
...   echo=True        # Whether to echo the prompt
... 

Error

CUDA error: the provided PTX was compiled with an unsupported toolchain.
  current device: 0, in function ggml_cuda_op_flatten at /tmp/pip-install-r5wzw2ak/llama-cpp-python_a072f9ea5ea14ad5807f3be7170295e9/vendor/llama.cpp/ggml-cuda.cu:9209
  cudaGetLastError()
GGML_ASSERT: /tmp/pip-install-r5wzw2ak/llama-cpp-python_a072f9ea5ea14ad5807f3be7170295e9/vendor/llama.cpp/ggml-cuda.cu:241: !"CUDA error"
No symbol table is loaded.  Use the "file" command.
ptrace: Operation not permitted.
No stack.
The program is not being run.
Aborted (core dumped)

Environment Details:

OS: Oracle Linux 7.9
GPU: NVIDIA A100
Python Version: Python 3.10
Transformers: transformers==4.37.2
LLAMA CPP PYTHON:  llama_cpp_python==0.2.44. (Installed using CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install llama-cpp-python)
NVIDIA SMI OUTPUT: $ nvidia-smi
Mon Feb 19 05:18:54 2024       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.129.03             Driver Version: 535.129.03   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA A100 80GB PCIe          On  | 00000000:25:00.0 Off |                    0 |
| N/A   32C    P0              64W / 300W |    848MiB / 81920MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA A100 80GB PCIe          On  | 00000000:81:00.0 Off |                    0 |
| N/A   29C    P0              43W / 300W |      7MiB / 81920MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+

Thanks in Advance!!

Sign up or log in to comment