Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

build failed on Mac / M1 #8

Open
alexng88 opened this issue Jul 22, 2023 · 3 comments
Open

build failed on Mac / M1 #8

alexng88 opened this issue Jul 22, 2023 · 3 comments

Comments

@alexng88
Copy link

Hello,
I follow the instruction and run pip install -r requirememts_mac.txt. and it shows the error

Building wheels for collected packages: xformers
Building wheel for xformers (setup.py) ... error
error: subprocess-exited-with-error

× python setup.py bdist_wheel did not run successfully.
│ exit code: 1
╰─> [216 lines of output]
running bdist_wheel
/Users/development/opt/anaconda3/envs/ollm/lib/python3.10/site-packages/torch/utils/cpp_extension.py:476: UserWarning: Attempted to use ninja as the BuildExtension backend but we could not find ninja.. Falling back to using the slow distutils backend.
warnings.warn(msg.format('we could not find ninja.'))
running build
running build_py
creating build
creating build/lib.macosx-10.9-x86_64-cpython-310
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers
copying xformers/version.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers
copying xformers/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers
copying xformers/test.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers
copying xformers/utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers
copying xformers/_cpp_lib.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers
copying xformers/info.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/fused_linear_layer.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/sum_strided.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/k_activations.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/k_layer_norm.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/k_sum.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/k_fused_matmul_fw.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/dropout.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/k_dropout.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/softmax.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/layer_norm.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/k_fused_matmul_bw.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
copying xformers/triton/k_softmax.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/triton
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/simplicial_embedding.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/residual.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/reversible.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/activations.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/multi_head_dispatch.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/input_projection.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
copying xformers/components/patch_embedding.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_mem_eff_attention.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_nvfuser.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_indexing.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_mlp.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_triton_stride_sum.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_blocksparse_transformers.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_transformer.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_revnet.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_swiglu.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_triton_layernorm.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_causal_blocksparse.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_triton_fused_linear.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_triton_blocksparse.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_triton_softmax.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_encoder.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_triton_dropout.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_pytorch_transformer.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_nystrom_utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_multi_head_dispatch.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_sddmm.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
copying xformers/benchmarks/benchmark_core.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops
copying xformers/ops/swiglu_op.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops
copying xformers/ops/unbind.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops
copying xformers/ops/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops
copying xformers/ops/common.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops
copying xformers/ops/indexing.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/profiler
copying xformers/profiler/device_limits.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/profiler
copying xformers/profiler/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/profiler
copying xformers/profiler/api.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/profiler
copying xformers/profiler/slow_ops_profiler.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/profiler
copying xformers/profiler/profiler.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/profiler
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/sparse
copying xformers/sparse/_csr_ops.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/sparse
copying xformers/sparse/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/sparse
copying xformers/sparse/utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/sparse
copying xformers/sparse/blocksparse_tensor.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/sparse
copying xformers/sparse/csr_tensor.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/sparse
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/helpers
copying xformers/helpers/test_utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/helpers
copying xformers/helpers/hierarchical_configs.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/helpers
copying xformers/helpers/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/helpers
copying xformers/helpers/timm_sparse_attention.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/helpers
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/fused_softmax.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/flash_blocksparse_attn_interface.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/flash_blocksparse_attention.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/bert_padding.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/flash_attn_triton_og.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/flash_attention.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/flash_attn_triton.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
copying xformers/_flash_attn/flash_attn_interface.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/factory
copying xformers/factory/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/factory
copying xformers/factory/hydra_helper.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/factory
copying xformers/factory/block_factory.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/factory
copying xformers/factory/model_factory.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/factory
copying xformers/factory/block_configs.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/factory
copying xformers/factory/weight_init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/factory
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/global_tokens.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/ortho.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/blocksparse.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/local.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/compositional.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/pooling.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/_sputnik_sparse.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/core.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/lambda_layer.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/random.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/fourier_mix.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/scaled_dot_product.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/attention_mask.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/linformer.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/attention_patterns.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/visual.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/sparsity_config.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/nystrom.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/favor.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
copying xformers/components/attention/base.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/feedforward
copying xformers/components/feedforward/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/feedforward
copying xformers/components/feedforward/mixture_of_experts.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/feedforward
copying xformers/components/feedforward/mlp.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/feedforward
copying xformers/components/feedforward/conv_mlp.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/feedforward
copying xformers/components/feedforward/fused_mlp.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/feedforward
copying xformers/components/feedforward/base.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/feedforward
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/positional_embedding
copying xformers/components/positional_embedding/vocab.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/positional_embedding
copying xformers/components/positional_embedding/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/positional_embedding
copying xformers/components/positional_embedding/param.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/positional_embedding
copying xformers/components/positional_embedding/sine.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/positional_embedding
copying xformers/components/positional_embedding/rotary.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/positional_embedding
copying xformers/components/positional_embedding/base.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/positional_embedding
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/nvfuser
copying xformers/components/nvfuser/bias_dropout_res_layernorm.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/nvfuser
copying xformers/components/nvfuser/bias_act_dropout.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/nvfuser
copying xformers/components/nvfuser/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/nvfuser
copying xformers/components/nvfuser/utils.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/nvfuser
copying xformers/components/nvfuser/bias_dropout_res.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/nvfuser
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention/feature_maps
copying xformers/components/attention/feature_maps/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention/feature_maps
copying xformers/components/attention/feature_maps/softmax.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention/feature_maps
copying xformers/components/attention/feature_maps/base.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/components/attention/feature_maps
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA
copying xformers/benchmarks/LRA/batch_submit.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA
copying xformers/benchmarks/LRA/batch_fetch_results.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA
copying xformers/benchmarks/LRA/run_with_submitit.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA
copying xformers/benchmarks/LRA/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA
copying xformers/benchmarks/LRA/run_tasks.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA
copying xformers/benchmarks/LRA/run_grid_search.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA/code
copying xformers/benchmarks/LRA/code/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA/code
copying xformers/benchmarks/LRA/code/model_wrapper.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA/code
copying xformers/benchmarks/LRA/code/dataset.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/benchmarks/LRA/code
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/triton.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/dispatch.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/attn_bias.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/common.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/flash.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/small_k.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
copying xformers/ops/fmha/cutlass.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/ops/fmha
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/losses
copying xformers/_flash_attn/losses/cross_entropy.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/losses
copying xformers/_flash_attn/losses/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/losses
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/layers
copying xformers/_flash_attn/layers/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/layers
copying xformers/_flash_attn/layers/patch_embed.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/layers
copying xformers/_flash_attn/layers/rotary.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/layers
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/utils
copying xformers/_flash_attn/utils/pretrained.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/utils
copying xformers/_flash_attn/utils/generation.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/utils
copying xformers/_flash_attn/utils/benchmark.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/utils
copying xformers/_flash_attn/utils/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/utils
copying xformers/_flash_attn/utils/distributed.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/utils
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/gptj.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/opt.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/llama.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/vit.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/bert.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/gpt_neox.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
copying xformers/_flash_attn/models/gpt.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/models
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/ops
copying xformers/_flash_attn/ops/activations.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/ops
copying xformers/_flash_attn/ops/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/ops
copying xformers/_flash_attn/ops/fused_dense.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/ops
copying xformers/_flash_attn/ops/rms_norm.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/ops
copying xformers/_flash_attn/ops/layer_norm.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/ops
creating build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/modules
copying xformers/_flash_attn/modules/embedding.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/modules
copying xformers/_flash_attn/modules/init.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/modules
copying xformers/_flash_attn/modules/mlp.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/modules
copying xformers/_flash_attn/modules/block.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/modules
copying xformers/_flash_attn/modules/mha.py -> build/lib.macosx-10.9-x86_64-cpython-310/xformers/_flash_attn/modules
running build_ext
building 'xformers._C' extension
creating build/temp.macosx-10.9-x86_64-cpython-310
creating build/temp.macosx-10.9-x86_64-cpython-310/xformers
creating build/temp.macosx-10.9-x86_64-cpython-310/xformers/csrc
creating build/temp.macosx-10.9-x86_64-cpython-310/xformers/csrc/attention
creating build/temp.macosx-10.9-x86_64-cpython-310/xformers/csrc/attention/autograd
creating build/temp.macosx-10.9-x86_64-cpython-310/xformers/csrc/attention/cpu
creating build/temp.macosx-10.9-x86_64-cpython-310/xformers/csrc/indexing
creating build/temp.macosx-10.9-x86_64-cpython-310/xformers/csrc/swiglu
clang -Wno-unused-result -Wsign-compare -Wunreachable-code -DNDEBUG -fwrapv -O2 -Wall -fPIC -O2 -isystem /Users/development/opt/anaconda3/envs/ollm/include -fPIC -O2 -isystem /Users/development/opt/anaconda3/envs/ollm/include -I/private/var/folders/0m/fj0bcx4n16dd2snclpw414080000gn/T/pip-install-9pmf8iz7/xformers_fec6780c30d34980ba5c4acf372654b4/xformers/csrc -I/Users/development/opt/anaconda3/envs/ollm/lib/python3.10/site-packages/torch/include -I/Users/development/opt/anaconda3/envs/ollm/lib/python3.10/site-packages/torch/include/torch/csrc/api/include -I/Users/development/opt/anaconda3/envs/ollm/lib/python3.10/site-packages/torch/include/TH -I/Users/development/opt/anaconda3/envs/ollm/lib/python3.10/site-packages/torch/include/THC -I/Users/development/opt/anaconda3/envs/ollm/include/python3.10 -c xformers/csrc/attention/attention.cpp -o build/temp.macosx-10.9-x86_64-cpython-310/xformers/csrc/attention/attention.o -O3 -fopenmp -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE="_clang" -DPYBIND11_STDLIB="_libcpp" -DPYBIND11_BUILD_ABI="_cxxabi1002" -DTORCH_EXTENSION_NAME=_C -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++17
clang: error: unsupported option '-fopenmp'
error: command '/usr/bin/clang' failed with exit code 1
[end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.
ERROR: Failed building wheel for xformers
Running setup.py clean for xformers
Failed to build xformers
ERROR: Could not build wheels for xformers, which is required to install pyproject.toml-based projects
@Uminosachi
Copy link
Owner

In order to build xformers on an M1 Mac, you need to install some additional libraries, namely libomp and llvm. You can use the brew package manager to do this.

If you haven't installed Homebrew yet, install it:

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install.sh)"
echo 'export PATH="/opt/homebrew/bin:$PATH"' >> ~/.zshrc
echo 'export PATH="/opt/homebrew/sbin:$PATH"' >> ~/.zshrc
source ~/.zshrc

Install libomp:

arch -arm64 brew install libomp
echo 'export PATH="/opt/homebrew/opt/libomp/bin:$PATH"' >> ~/.zshrc
source ~/.zshrc

Install llvm:

arch -arm64 brew install llvm
echo 'export PATH="/opt/homebrew/opt/llvm/bin:$PATH"' >> ~/.zshrc
source ~/.zshrc

@alexng88
Copy link
Author

thanks it works!
How can we download another model? for example if i want to add llama2 how can we add?

thanks

@Uminosachi
Copy link
Owner

I am considering adding Llama2 to the list in ollm_model_ids within ollm_app.py. However, to download the Llama2 model, one needs to obtain prior access from Meta AI, so this might take some time.

You can try Llama2 on the following demo site:

https://huggingface.co/spaces/ysharma/Explore_llamav2_with_TGI

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@alexng88 @Uminosachi