I encountered an error when creating a small-rf model. I tried uninstalling FlashAttention, but it didn't work, and I couldn't change it to another usage method. log: OS windows11 GPU:gtx1070 RAM 32gb VRAM 8gb theDAW UI initialized x2 API healthy, model loading x2 XR control bus connected. Loaded 0 entries from backend-local [+0.0s] CREATE pressed: model=small-rf duration=56.5s steps=14 seed=-1 prompt="120pbm" [+0.0s] POST /api/generate-jobs — model=small-rf duration=56.5s steps=14 seed=-1 [+35.4s] POST /api/generate-jobs → 200 OK — job_id=d805cea3 (server received the job) [+35.4s] Job queued: d805cea3 — waiting for backend to start sampling [+35.7s] Job running: d805cea3 — sampler started (14 steps requested) FlashAttention only supports Ampere GPUs or newer. Exception raised from mha_varlen_fwd at D:\a\flash-attention\flash-attention\csrc\flash_attn\flash_api.cpp:542 (most recent call first): 00007FFFAFAD220900007FFFAFAD2160 c10.dll!c10::Error::Error [ @ ] 00007FFFAFAD0C0A00007FFFAFAD0BB0 c10.dll!c10::detail::torchCheckFail [ @ ] 00007FFC954A46E500007FFC95495BB0 flash_attn_2_cuda.cp310-win_amd64.pyd!c10::ivalue::Object::operator= [ @ ] 00007FFC954B791900007FFC954AC320 flash_attn_2_cuda.cp310-win_amd64.pyd!PyInit_flash_attn_2_cuda [ @ ] 00007FFC954B5A1F00007FFC954AC320 flash_attn_2_cuda.cp310-win_amd64.pyd!PyInit_flash_attn_2_cuda [ @ ] 00007FFC954B5AF400007FFC954AC320 flash_attn_2_cuda.cp310-win_amd64.pyd!PyInit_flash_attn_2_cuda [ @ ] 00007FFC9549E64500007FFC95495BB0 flash_attn_2_cuda.cp310-win_amd64.pyd!c10::ivalue::Object::operator= [ @ ] 00007FFFB1DDBF6E00007FFFB1DDBA8C python310.dll!PyList_New [ @ ] 00007FFFB1E196C000007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1E479A400007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ] 00007FFFB1E4787D00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E1B52800007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1E479A400007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ] 00007FFFB1E4787D00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E1B52800007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1E479A400007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ] 00007FFFB1E4787D00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E1B52800007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1E479A400007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ] 00007FFFB1E4787D00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E1B52800007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1E479A400007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ] 00007FFFB1E4777F00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFE5BCA598200007FFE5BC7F6C0 torch_python.dll!torch::FunctionParameter::type_name [ @ ] 00007FFD0AF1CDEC00007FFD0AF1CC80 torch_cpu.dll!c10::Dispatcher::callBoxed [ @ ] 00007FFE5BA7B3E000007FFE5BA7B330 torch_python.dll!torch::jit::invokeOperatorFromPython [ @ ] 00007FFE5BA7846700007FFE5BA78320 torch_python.dll!torch::jit::_get_operation_for_overload_or_packet [ @ ] 00007FFE5B9E1B7600007FFE5B89BF20 torch_python.dll!registerPythonTensorClass [ @ ] 00007FFE5B9881C600007FFE5B89BF20 torch_python.dll!registerPythonTensorClass [ @ ] 00007FFE5B3D156B00007FFE5B3CFFF0 torch_python.dll!c10::ivalue::Future::devices [ @ ] 00007FFFB1DDBF6E00007FFFB1DDBA8C python310.dll!PyList_New [ @ ] 00007FFFB1E477C400007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E478BF00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E1B52800007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1DEF4AC00007FFFB1DEF3C4 python310.dll!PyObject_FastCallDictTstate [ @ ] 00007FFFB1EC863300007FFFB1EC85B4 python310.dll!PyObject_Call_Prepend [ @ ] 00007FFFB1EC856000007FFFB1EC817C python310.dll!PyObject_SelfIter [ @ ] 00007FFFB1DF040E00007FFFB1DF02C0 python310.dll!PyObject_MakeTpCall [ @ ] 00007FFFB1E1BDDB00007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1E479A400007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ] 00007FFFB1E4777F00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFE5B87228100007FFE5B86CDF0 torch_python.dll!torch::autograd::registerFunctionPreHook [ @ ] 00007FFFB1DDBF9D00007FFFB1DDBA8C python310.dll!PyList_New [ @ ] 00007FFFB1E477C400007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E478BF00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E1B52800007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E12C9E00007FFFB1E11970 python310.dll!PyObject_GC_Del [ @ ] 00007FFFB1E1719200007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E147DD00007FFFB1E14750 python310.dll!PyFunction_Vectorcall [ @ ] 00007FFFB1E1B78500007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E1856400007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E12C9E00007FFFB1E11970 python310.dll!PyObject_GC_Del [ @ ] 00007FFFB1E47A0000007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ] 00007FFFB1E4787D00007FFFB1E47730 python310.dll!PyObject_Call [ @ ] 00007FFFB1E1B52800007FFFB1E15760 python310.dll!PyEval_EvalFrameDefault [ @ ] 00007FFFB1E12C9E00007FFFB1E11970 python310.dll!PyObject_GC_Del [ @ ] 00007FFFB1E47A0000007FFFB1E47948 python310.dll!PyVectorcall_Call [ @ ]

Fail launch llmfit from pinokio

FlashAttention error

Mac M3 48Gb RAM

Radar

Popular builders