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

Quantize before import and conversion to Linalg during CompiledModule.__new__() #211

Closed
wants to merge 1 commit into from
Closed

Quantize before import and conversion to Linalg during CompiledModule.__new__() #211

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions python/shark_turbine/aot/compiled_module.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,6 +45,8 @@
"CompiledModule",
]

from shark_turbine.transforms.rewriter import Pass # for type annotations

################################################################################
# Data structures
################################################################################
Expand Down Expand Up @@ -472,6 +474,7 @@ def __new__(
context: Optional[Context] = None,
module_op: Optional[Operation] = None,
import_to: Union[ImportPhase, None, str] = "full",
pre_import_passes=List[Pass]
):
import_to = ImportPhase.parse(import_to)
self = super().__new__(cls)
Expand Down Expand Up @@ -538,6 +541,13 @@ def invoke_with_self(*args, **kwargs):
do_export(proc_def)

module_builder.finalize_construct()

# `run_import` transforms module from torch to linalg, and passes like MMGroupQuantRewriterPass need to be run before
module_op = CompiledModule.get_mlir_module(self)
from shark_turbine.transforms.quantization.mm_group_quant import MMGroupQuantRewriterPass
for p in pre_import_passes:
p(module_op).run()

CompiledModule.run_import(self, import_to)
return self

Expand Down
10 changes: 3 additions & 7 deletions python/turbine_models/custom_models/stateless_llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -197,15 +197,11 @@ def forward(token0: torch.Tensor, *state0_flat):
return token1, *state1_flat

import_to = "INPUT" if compile_to == "linalg" else "IMPORT"
inst = StateUpdateModule(context=Context(), import_to=import_to)
# TODO: Integrate with external parameters to actually be able to run
# TODO: Make more generalizable to be able to quantize with all compile_to options
pre_import_passes = []
if quantization == "int4" and not compile_to == "linalg":
from shark_turbine.transforms.quantization import mm_group_quant

mm_group_quant.MMGroupQuantRewriterPass(
CompiledModule.get_mlir_module(inst).operation
).run()
pre_import_passes.append(mm_group_quant.MMGroupQuantRewriterPass)
inst = StateUpdateModule(context=Context(), import_to=import_to, pre_import_passes=pre_import_passes)
module_str = str(CompiledModule.get_mlir_module(inst))
safe_name = hf_model_name.split("/")[-1].strip()
safe_name = re.sub("-", "_", safe_name)
Expand Down
Loading