Using `is_flash_attn_available` is deprecated and will be removed in v4.38. Please use `is_flash_attn_2_available` instead.
Traceback (most recent call last):
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/tools/gradio/app.py", line 10, in <module>
from levo_inference_lowmem import LeVoInference
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/tools/gradio/levo_inference_lowmem.py", line 11, in <module>
from codeclm.trainer.codec_song_pl import CodecLM_PL
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/codeclm/trainer/codec_song_pl.py", line 18, in <module>
from codeclm.models import builders
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/codeclm/models/__init__.py", line 11, in <module>
from .codeclm import CodecLM
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/codeclm/models/codeclm.py", line 12, in <module>
from .lm_levo import LmModel
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/codeclm/models/lm_levo.py", line 10, in <module>
from codeclm.models.levo import CausalLM, LlamaConfig
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/codeclm/models/levo.py", line 2, in <module>
from .llama.modeling_llama import LlamaConfig, CausalLMOutputWithPast, BaseModelOutputWithPast, LlamaDecoderLayer, LlamaRMSNorm
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/codeclm/models/llama/modeling_llama.py", line 45, in <module>
from flash_attn import flash_attn_func, flash_attn_varlen_func
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/env/lib/python3.10/site-packages/flash_attn/__init__.py", line 3, in <module>
from flash_attn.flash_attn_interface import (
File "/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/env/lib/python3.10/site-packages/flash_attn/flash_attn_interface.py", line 15, in <module>
import flash_attn_2_cuda as flash_attn_gpu
ImportError: /media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app/env/lib/python3.10/site-packages/flash_attn_2_cuda.cpython-310-x86_64-linux-gnu.so: undefined symbol: _ZN3c105ErrorC2ENS_14SourceLocationENSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE
(env) (base) marius@zbook:/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app$ python -c "import torch; print(torch.__version__)"
2.6.0+cu124
(env) (base) marius@zbook:/media/marius/FIVEH1/PINOCHIO/api/SongGeneration.git/app$ pip install flash-attn --no-build-isolation
Requirement already satisfied: flash-attn in ./env/lib/python3.10/site-packages (2.7.4.post1)
Requirement already satisfied: torch in ./env/lib/python3.10/site-packages (from flash-attn) (2.6.0)
Requirement already satisfied: einops in ./env/lib/python3.10/site-packages (from flash-attn) (0.8.2)
Requirement already satisfied: filelock in ./env/lib/python3.10/site-packages (from torch->flash-attn) (3.24.3)
Requirement already satisfied: typing-extensions>=4.10.0 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (4.15.0)
Requirement already satisfied: networkx in ./env/lib/python3.10/site-packages (from torch->flash-attn) (3.4.2)
Requirement already satisfied: jinja2 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (3.1.6)
Requirement already satisfied: fsspec in ./env/lib/python3.10/site-packages (from torch->flash-attn) (2026.2.0)
Requirement already satisfied: nvidia-cuda-nvrtc-cu12==12.4.127 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (12.4.127)
Requirement already satisfied: nvidia-cuda-runtime-cu12==12.4.127 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (12.4.127)
Requirement already satisfied: nvidia-cuda-cupti-cu12==12.4.127 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (12.4.127)
Requirement already satisfied: nvidia-cudnn-cu12==9.1.0.70 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (9.1.0.70)
Requirement already satisfied: nvidia-cublas-cu12==12.4.5.8 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (12.4.5.8)
Requirement already satisfied: nvidia-cufft-cu12==11.2.1.3 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (11.2.1.3)
Requirement already satisfied: nvidia-curand-cu12==10.3.5.147 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (10.3.5.147)
Requirement already satisfied: nvidia-cusolver-cu12==11.6.1.9 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (11.6.1.9)
Requirement already satisfied: nvidia-cusparse-cu12==12.3.1.170 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (12.3.1.170)
Requirement already satisfied: nvidia-cusparselt-cu12==0.6.2 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (0.6.2)
Requirement already satisfied: nvidia-nccl-cu12==2.21.5 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (2.21.5)
Requirement already satisfied: nvidia-nvtx-cu12==12.4.127 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (12.4.127)
Requirement already satisfied: nvidia-nvjitlink-cu12==12.4.127 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (12.4.127)
Requirement already satisfied: triton==3.2.0 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (3.2.0)
Requirement already satisfied: sympy==1.13.1 in ./env/lib/python3.10/site-packages (from torch->flash-attn) (1.13.1)
Requirement already satisfied: mpmath<1.4,>=1.1.0 in ./env/lib/python3.10/site-packages (from sympy==1.13.1->torch->flash-attn) (1.3.0)
Requirement already satisfied: MarkupSafe>=2.0 in ./env/lib/python3.10/site-packages (from jinja2->torch->flash-attn) (3.0.3)
Is there a work around this
python -c "import torch; print(torch.version)"
2.6.0+cu124
pip install -r ./requirements.txt
all good, but....
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2025 NVIDIA Corporation
Built on Tue_Dec_16_07:23:41_PM_PST_2025
Cuda compilation tools, release 13.1, V13.1.115
Build cuda_13.1.r13.1/compiler.37061995_0