Remove datapipe from dgl dependency. (#7667)

Co-authored-by: Ubuntu <[email protected]>

Author: frozenbugs

docker/install/conda_env/torch_cpu_pip.txt

@@ -18,7 +18,6 @@ requests[security]==2.28
 scikit-learn
 scipy
 torch==2.3.0+cpu
-torchdata
 torcheval
 torchmetrics
 torch_geometric

docker/install/conda_env/torch_gpu_pip.txt

@@ -16,7 +16,6 @@ requests[security]==2.28
 scikit-learn
 scipy
 torch==2.3.0+cu121
-torchdata
 torcheval
 torchmetrics
 torch_geometric

docs/source/guide/minibatch-parallelism.rst

@@ -24,7 +24,7 @@ generate a minibatch, including:
     dataloader = gb.DataLoader(datapipe)
 
 All these stages are implemented in separate
-`IterableDataPipe <https://pytorch.org/data/main/torchdata.datapipes.iter.html>`__
+`IterableDataPipe <https://pytorch.org/data/0.7/torchdata.datapipes.iter.html>`__
 and stacked together with `PyTorch DataLoader <https://pytorch.org/docs/stable/data
 .html#torch.utils.data.DataLoader>`__.
 This design allows us to easily customize the data loading process by
@@ -44,7 +44,7 @@ data movement overheads between processes.
 
 What's more, in order to overlap the data movement and model computation, we
 wrap data pipes before ``copy_to`` with
-`torchdata.datapipes.iter.Perfetcher <https://pytorch.org/data/main/generated/
+`torchdata.datapipes.iter.Perfetcher <https://pytorch.org/data/0.7/generated/
 torchdata.datapipes.iter.Prefetcher.html>`__
 which prefetches elements from previous data pipes and puts them into a buffer.
 Such prefetching is totally transparent to users and requires no extra code. It

python/setup.py

@@ -226,14 +226,11 @@ def get_lib_file_path(lib_name, backend=""):
     "requests>=2.19.0",
     "tqdm",
     "psutil>=5.8.0",
-    "torchdata>=0.5.0",
     "pandas",
     "packaging",
     "pyyaml",
     "pydantic>=2.0",
 ]
-if "DGLBACKEND" in os.environ and os.environ["DGLBACKEND"] != "pytorch":
-    install_requires.pop(install_requires.index("torchdata>=0.5.0"))
 
 setup(
     name="dgl" + os.getenv("DGL_PACKAGE_SUFFIX", ""),

script/dgl_dev.yml.template

@@ -31,7 +31,6 @@ dependencies:
     - scikit-learn
     - scipy
     - torch==__TORCH_VERSION__
-    - torchdata>=0.5.0
     - torcheval
     - torchmetrics
     - torch_geometric

tests/lint/pylintrc

@@ -213,7 +213,7 @@ function-naming-style=snake_case
 # op - operators
 # ty - type
 # A, B, C, W - for tensor operators like matmul
-# dp - DataPipes (see https://pytorch.org/data/main/torchdata.datapipes.iter.html)
+# dp - DataPipes (see https://pytorch.org/data/0.7/torchdata.datapipes.iter.html)
 # it - iterators
 good-names=f,i,j,k,u,v,e,n,m,w,x,y,z,s,d,t,r,g,G,hg,sg,fn,ex,Run,_,us,vs,gs,es,op,ty,A,B,C,W,a,b,N,D1,D2,R,dp,it