Add op ir #270

weilinquan · 2023-12-31T03:27:38Z

You can also run llama by BuddyLlama.

[frontend] refine implementation of `op_def`

zhanghb97

Failed Tests (13):
  BUDDY :: Python/test_addmm.py
  BUDDY :: Python/test_amax.py
  BUDDY :: Python/test_arith_sub.py
  BUDDY :: Python/test_convert_element_type.py
  BUDDY :: Python/test_embedding.py
  BUDDY :: Python/test_exp.py
  BUDDY :: Python/test_expand.py
  BUDDY :: Python/test_permute.py
  BUDDY :: Python/test_reshape.py
  BUDDY :: Python/test_select.py
  BUDDY :: Python/test_sum.py
  BUDDY :: Python/test_tanh.py
  BUDDY :: Python/test_var_mean.py

I encountered above failed when I run the ninja check-buddy. Please have a check.

weilinquan · 2024-01-02T06:39:41Z

Failed Tests (13):
  BUDDY :: Python/test_addmm.py
  BUDDY :: Python/test_amax.py
  BUDDY :: Python/test_arith_sub.py
  BUDDY :: Python/test_convert_element_type.py
  BUDDY :: Python/test_embedding.py
  BUDDY :: Python/test_exp.py
  BUDDY :: Python/test_expand.py
  BUDDY :: Python/test_permute.py
  BUDDY :: Python/test_reshape.py
  BUDDY :: Python/test_select.py
  BUDDY :: Python/test_sum.py
  BUDDY :: Python/test_tanh.py
  BUDDY :: Python/test_var_mean.py

I encountered above failed when I run the ninja check-buddy. Please have a check.

Thank you for your comment. This because tosa's op set have not adapted for current graph IR. We can add tosa support later in this PR.

…an run on new graph ir frontend now.

[frontend] Add tosa support for graph ir

zhanghb97 · 2024-01-04T12:36:38Z

Failed to build LLaMA example:

[1/5] Generating llama.mlir and arg0.data...
FAILED: ../examples/BuddyLlama/llama.mlir ../examples/BuddyLlama/arg0.data /home/zhb/buddy-mlir/examples/BuddyLlama/llama.mlir /home/zhb/buddy-mlir/examples/BuddyLlama/arg0.data 
cd /home/zhb/buddy-mlir/build/examples/BuddyLlama && /home/zhb/miniforge3/envs/test/bin/python3 /home/zhb/buddy-mlir/examples/BuddyLlama/import-llama2.py
You are using the default legacy behaviour of the <class 'transformers.models.llama.tokenization_llama.LlamaTokenizer'>. If you see this, DO NOT PANIC! This is expected, and simply means that the `legacy` (previous) behavior will be used so nothing changes for you. If you want to use the new behaviour, set `legacy=False`. This should only be set if you understand what it means, and thouroughly read the reason why this was added as explained in https://github.com/huggingface/transformers/pull/24565
Loading checkpoint shards: 100%|██████████| 2/2 [00:31<00:00, 15.60s/it]
Traceback (most recent call last):
  File "/home/zhb/buddy-mlir/examples/BuddyLlama/import-llama2.py", line 54, in <module>
    graphs = dynamo_compiler.importer(model, data)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/frontend.py", line 223, in importer
    model_opt(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 328, in _fn
    return fn(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 490, in catch_errors
    return callback(frame, cache_entry, hooks, frame_state)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 641, in _convert_frame
    result = inner_convert(frame, cache_size, hooks, frame_state)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 133, in _fn
    return fn(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 389, in _convert_frame_assert
    return _compile(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 569, in _compile
    guarded_code = compile_inner(code, one_graph, hooks, transform)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/utils.py", line 189, in time_wrapper
    r = func(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 491, in compile_inner
    out_code = transform_code_object(code, transform)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/bytecode_transformation.py", line 1028, in transform_code_object
    transformations(instructions, code_options)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 458, in transform
    tracer.run()
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 2069, in run
    super().run()
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 719, in run
    and self.step()
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 683, in step
    getattr(self, inst.opname)(inst)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 2157, in RETURN_VALUE
    self.output.compile_subgraph(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 857, in compile_subgraph
    self.compile_and_call_fx_graph(tx, pass2.graph_output_vars(), root)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/contextlib.py", line 79, in inner
    return func(*args, **kwds)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 957, in compile_and_call_fx_graph
    compiled_fn = self.call_user_compiler(gm)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/utils.py", line 189, in time_wrapper
    r = func(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 1024, in call_user_compiler
    raise BackendCompilerFailed(self.compiler_fn, e).with_traceback(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 1009, in call_user_compiler
    compiled_fn = compiler_fn(gm, self.example_inputs())
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/repro/after_dynamo.py", line 117, in debug_wrapper
    compiled_gm = compiler_fn(gm, example_inputs)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/frontend.py", line 188, in _compile_fx
    return aot_module_simplified(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 3891, in aot_module_simplified
    compiled_fn = create_aot_dispatcher_function(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/utils.py", line 189, in time_wrapper
    r = func(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 3429, in create_aot_dispatcher_function
    compiled_fn = compiler_fn(flat_fn, fake_flat_args, aot_config, fw_metadata=fw_metadata)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 2212, in aot_wrapper_dedupe
    return compiler_fn(flat_fn, leaf_flat_args, aot_config, fw_metadata=fw_metadata)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 2392, in aot_wrapper_synthetic_base
    return compiler_fn(flat_fn, flat_args, aot_config, fw_metadata=fw_metadata)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 1573, in aot_dispatch_base
    compiled_fw = compiler(fw_module, flat_args)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/frontend.py", line 181, in _compiler
    return graph.dynamo_run()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 230, in dynamo_run
    self.compile()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 223, in compile
    self.lower_to_top_level_ir()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 140, in lower_to_top_level_ir
    self._imported_module = fx_importer.import_graph()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 392, in import_graph
    def generated_func(*args):
  File "/home/zhb/buddy-mlir/llvm/build/tools/mlir/python_packages/mlir_core/mlir/dialects/func.py", line 195, in decorator
    return_values = f(*func_args, **func_kwargs)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 416, in generated_func
    self._import_op(node)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 472, in _import_op
    self._ops_registry[op_name](node, self._symbol_table)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/ops/tosa.py", line 873, in transpose_op
    assert len(node.args) == 3, "Input tensor must be 3D"
torch._dynamo.exc.BackendCompilerFailed: backend='_compile_fx' raised:
AssertionError: Input tensor must be 3D

Set TORCH_LOGS="+dynamo" and TORCHDYNAMO_VERBOSE=1 for more information


You can suppress this exception and fall back to eager by setting:
    import torch._dynamo
    torch._dynamo.config.suppress_errors = True

ninja: build stopped: subcommand failed.

weilinquan · 2024-01-04T13:35:51Z

Failed to build LLaMA example:

[1/5] Generating llama.mlir and arg0.data...
FAILED: ../examples/BuddyLlama/llama.mlir ../examples/BuddyLlama/arg0.data /home/zhb/buddy-mlir/examples/BuddyLlama/llama.mlir /home/zhb/buddy-mlir/examples/BuddyLlama/arg0.data 
cd /home/zhb/buddy-mlir/build/examples/BuddyLlama && /home/zhb/miniforge3/envs/test/bin/python3 /home/zhb/buddy-mlir/examples/BuddyLlama/import-llama2.py
You are using the default legacy behaviour of the <class 'transformers.models.llama.tokenization_llama.LlamaTokenizer'>. If you see this, DO NOT PANIC! This is expected, and simply means that the `legacy` (previous) behavior will be used so nothing changes for you. If you want to use the new behaviour, set `legacy=False`. This should only be set if you understand what it means, and thouroughly read the reason why this was added as explained in https://github.com/huggingface/transformers/pull/24565
Loading checkpoint shards: 100%|██████████| 2/2 [00:31<00:00, 15.60s/it]
Traceback (most recent call last):
  File "/home/zhb/buddy-mlir/examples/BuddyLlama/import-llama2.py", line 54, in <module>
    graphs = dynamo_compiler.importer(model, data)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/frontend.py", line 223, in importer
    model_opt(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 328, in _fn
    return fn(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 490, in catch_errors
    return callback(frame, cache_entry, hooks, frame_state)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 641, in _convert_frame
    result = inner_convert(frame, cache_size, hooks, frame_state)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 133, in _fn
    return fn(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 389, in _convert_frame_assert
    return _compile(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 569, in _compile
    guarded_code = compile_inner(code, one_graph, hooks, transform)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/utils.py", line 189, in time_wrapper
    r = func(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 491, in compile_inner
    out_code = transform_code_object(code, transform)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/bytecode_transformation.py", line 1028, in transform_code_object
    transformations(instructions, code_options)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/convert_frame.py", line 458, in transform
    tracer.run()
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 2069, in run
    super().run()
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 719, in run
    and self.step()
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 683, in step
    getattr(self, inst.opname)(inst)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/symbolic_convert.py", line 2157, in RETURN_VALUE
    self.output.compile_subgraph(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 857, in compile_subgraph
    self.compile_and_call_fx_graph(tx, pass2.graph_output_vars(), root)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/contextlib.py", line 79, in inner
    return func(*args, **kwds)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 957, in compile_and_call_fx_graph
    compiled_fn = self.call_user_compiler(gm)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/utils.py", line 189, in time_wrapper
    r = func(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 1024, in call_user_compiler
    raise BackendCompilerFailed(self.compiler_fn, e).with_traceback(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/output_graph.py", line 1009, in call_user_compiler
    compiled_fn = compiler_fn(gm, self.example_inputs())
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/repro/after_dynamo.py", line 117, in debug_wrapper
    compiled_gm = compiler_fn(gm, example_inputs)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/frontend.py", line 188, in _compile_fx
    return aot_module_simplified(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 3891, in aot_module_simplified
    compiled_fn = create_aot_dispatcher_function(
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_dynamo/utils.py", line 189, in time_wrapper
    r = func(*args, **kwargs)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 3429, in create_aot_dispatcher_function
    compiled_fn = compiler_fn(flat_fn, fake_flat_args, aot_config, fw_metadata=fw_metadata)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 2212, in aot_wrapper_dedupe
    return compiler_fn(flat_fn, leaf_flat_args, aot_config, fw_metadata=fw_metadata)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 2392, in aot_wrapper_synthetic_base
    return compiler_fn(flat_fn, flat_args, aot_config, fw_metadata=fw_metadata)
  File "/home/zhb/miniforge3/envs/test/lib/python3.10/site-packages/torch/_functorch/aot_autograd.py", line 1573, in aot_dispatch_base
    compiled_fw = compiler(fw_module, flat_args)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/frontend.py", line 181, in _compiler
    return graph.dynamo_run()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 230, in dynamo_run
    self.compile()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 223, in compile
    self.lower_to_top_level_ir()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 140, in lower_to_top_level_ir
    self._imported_module = fx_importer.import_graph()
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 392, in import_graph
    def generated_func(*args):
  File "/home/zhb/buddy-mlir/llvm/build/tools/mlir/python_packages/mlir_core/mlir/dialects/func.py", line 195, in decorator
    return_values = f(*func_args, **func_kwargs)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 416, in generated_func
    self._import_op(node)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/graph/graph.py", line 472, in _import_op
    self._ops_registry[op_name](node, self._symbol_table)
  File "/home/zhb/buddy-mlir/build/python_packages/buddy/compiler/ops/tosa.py", line 873, in transpose_op
    assert len(node.args) == 3, "Input tensor must be 3D"
torch._dynamo.exc.BackendCompilerFailed: backend='_compile_fx' raised:
AssertionError: Input tensor must be 3D

Set TORCH_LOGS="+dynamo" and TORCHDYNAMO_VERBOSE=1 for more information


You can suppress this exception and fall back to eager by setting:
    import torch._dynamo
    torch._dynamo.config.suppress_errors = True

ninja: build stopped: subcommand failed.

@zhanghb97 I have fix this error. Please review it again. Thank you !

zhanghb97 · 2024-01-04T13:55:19Z

ninja check-buddy failed:

******************** TEST 'BUDDY :: Python/test_transpose.py' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
RUN: at line 1: /home/zhb/miniforge3/envs/test/bin/python3 /home/zhb/buddy-mlir/tests/Python/test_transpose.py 2>&1 | /home/zhb/buddy-mlir/llvm/build/bin/FileCheck /home/zhb/buddy-mlir/tests/Python/test_transpose.py
+ /home/zhb/buddy-mlir/llvm/build/bin/FileCheck /home/zhb/buddy-mlir/tests/Python/test_transpose.py
+ /home/zhb/miniforge3/envs/test/bin/python3 /home/zhb/buddy-mlir/tests/Python/test_transpose.py
/home/zhb/buddy-mlir/tests/Python/test_transpose.py:33:10: error: CHECK: expected string not found in input
# CHECK: %{{.*}} = tensor.empty()
         ^
<stdin>:2:20: note: scanning from here
 func.func @forward(%arg0: tensor<13x13x13xf32>) -> tensor<13x13x13xf32> {
                   ^
<stdin>:3:69: note: possible intended match here
 %0 = "tosa.const"() <{value = dense<[0, 2, 1]> : tensor<3xi32>}> : () -> tensor<3xi32>
                                                                    ^

Input file: <stdin>
Check file: /home/zhb/buddy-mlir/tests/Python/test_transpose.py

-dump-input=help explains the following input dump.

Input was:
<<<<<<
            1: module { 
            2:  func.func @forward(%arg0: tensor<13x13x13xf32>) -> tensor<13x13x13xf32> { 
check:33'0                        X~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ error: no match found
            3:  %0 = "tosa.const"() <{value = dense<[0, 2, 1]> : tensor<3xi32>}> : () -> tensor<3xi32> 
check:33'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
check:33'1                                                                         ?                    possible intended match
            4:  %1 = tosa.transpose %arg0, %0 : (tensor<13x13x13xf32>, tensor<3xi32>) -> tensor<13x13x13xf32> 
check:33'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
            5:  return %1 : tensor<13x13x13xf32> 
check:33'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
            6:  } 
check:33'0     ~~~
            7: } 
check:33'0     ~~
            8:  
check:33'0     ~
>>>>>>

--

********************
********************
Failed Tests (1):
  BUDDY :: Python/test_transpose.py

weilinquan · 2024-01-04T14:31:37Z

ninja check-buddy failed:

******************** TEST 'BUDDY :: Python/test_transpose.py' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
RUN: at line 1: /home/zhb/miniforge3/envs/test/bin/python3 /home/zhb/buddy-mlir/tests/Python/test_transpose.py 2>&1 | /home/zhb/buddy-mlir/llvm/build/bin/FileCheck /home/zhb/buddy-mlir/tests/Python/test_transpose.py
+ /home/zhb/buddy-mlir/llvm/build/bin/FileCheck /home/zhb/buddy-mlir/tests/Python/test_transpose.py
+ /home/zhb/miniforge3/envs/test/bin/python3 /home/zhb/buddy-mlir/tests/Python/test_transpose.py
/home/zhb/buddy-mlir/tests/Python/test_transpose.py:33:10: error: CHECK: expected string not found in input
# CHECK: %{{.*}} = tensor.empty()
         ^
<stdin>:2:20: note: scanning from here
 func.func @forward(%arg0: tensor<13x13x13xf32>) -> tensor<13x13x13xf32> {
                   ^
<stdin>:3:69: note: possible intended match here
 %0 = "tosa.const"() <{value = dense<[0, 2, 1]> : tensor<3xi32>}> : () -> tensor<3xi32>
                                                                    ^

Input file: <stdin>
Check file: /home/zhb/buddy-mlir/tests/Python/test_transpose.py

-dump-input=help explains the following input dump.

Input was:
<<<<<<
            1: module { 
            2:  func.func @forward(%arg0: tensor<13x13x13xf32>) -> tensor<13x13x13xf32> { 
check:33'0                        X~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ error: no match found
            3:  %0 = "tosa.const"() <{value = dense<[0, 2, 1]> : tensor<3xi32>}> : () -> tensor<3xi32> 
check:33'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
check:33'1                                                                         ?                    possible intended match
            4:  %1 = tosa.transpose %arg0, %0 : (tensor<13x13x13xf32>, tensor<3xi32>) -> tensor<13x13x13xf32> 
check:33'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
            5:  return %1 : tensor<13x13x13xf32> 
check:33'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
            6:  } 
check:33'0     ~~~
            7: } 
check:33'0     ~~
            8:  
check:33'0     ~
>>>>>>

--

********************
********************
Failed Tests (1):
  BUDDY :: Python/test_transpose.py

@zhanghb97 Sorry about that. I have fix it and run ninja check-buddy on my machine. I found it will output this.
******************** TEST 'BUDDY-EXAMPLES :: BuddyBert/bert.mlir' FAILED ********************
Test has no 'RUN:' line

Unresolved Tests (1):
BUDDY-EXAMPLES :: BuddyBert/bert.mlir

Testing Time: 0.47s

Total Discovered Tests: 88
Passed : 87 (98.86%)
Unresolved: 1 (1.14%)
FAILED: examples/CMakeFiles/check-examples /root/buddy-mlir/build/examples/CMakeFiles/check-examples
cd /root/buddy-mlir/build/examples && /root/anaconda3/bin/python /root/buddy-mlir/llvm/build/./bin/llvm-lit -sv /root/buddy-mlir/build/examples
It seems BuddyBert has not added to lit. Here should not have other errors.

zhanghb97 · 2024-01-04T14:39:10Z

@weilinquan No worries. Fixed the bert test: fc2948a

weilinquan · 2024-01-04T15:02:38Z

@weilinquan No worries. Fixed the bert test: fc2948a

@zhanghb97 Thank you for your comment ! I have add it and now ninja check-buddy is ok on my machine.

zhanghb97 · 2024-01-04T15:23:02Z

@weilinquan @xTayEx Thank you! Now, it works well. I am still reviewing the details.

weilinquan and others added 14 commits December 11, 2023 08:11

add graph lower

059c946

add jit support

811d8ec

add annotation

e556e96

add 10 ops

799fc47

add 25 ops

34aa136

add all ops

66378d9

fix lower bugs

a3e26e2

finish op lower debug

508ec7d

file format

5ccc08b

merge buddy main

c393614

fix buddygraph

c100e1f

fix buddygraph

0d7654e

[frontend] refine implementation of op_def

dfae011

Merge pull request #1 from xTayEx/add-ir-op

f2ef534

[frontend] refine implementation of `op_def`

zhanghb97 reviewed Jan 2, 2024

View reviewed changes

weilinquan and others added 7 commits January 2, 2024 09:06

change softmax lower

b6e812b

change softmax test

2e4f822

fix batch matmul

7b29807

[frontend] Add tosa support for graph ir

4d8b7eb

[frontend] add support for all current tosa operators. BERT example c…

3a781da

…an run on new graph ir frontend now.

[frontend] update tosa operator test

dfa69a2

Merge pull request #2 from xTayEx/xtayex_add-op-ir

67c9d40

[frontend] Add tosa support for graph ir

fix llama error

b9019fd

fix test_transpose error

023c1ea

add bert for lit

c8193e6

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add op ir #270

Add op ir #270

weilinquan commented Dec 31, 2023

zhanghb97 left a comment

weilinquan commented Jan 2, 2024

zhanghb97 commented Jan 4, 2024

weilinquan commented Jan 4, 2024

zhanghb97 commented Jan 4, 2024

weilinquan commented Jan 4, 2024 •

edited

Loading

zhanghb97 commented Jan 4, 2024

weilinquan commented Jan 4, 2024

zhanghb97 commented Jan 4, 2024

Add op ir #270

Are you sure you want to change the base?

Add op ir #270

Conversation

weilinquan commented Dec 31, 2023

zhanghb97 left a comment

Choose a reason for hiding this comment

weilinquan commented Jan 2, 2024

zhanghb97 commented Jan 4, 2024

weilinquan commented Jan 4, 2024

zhanghb97 commented Jan 4, 2024

weilinquan commented Jan 4, 2024 • edited Loading

zhanghb97 commented Jan 4, 2024

weilinquan commented Jan 4, 2024

zhanghb97 commented Jan 4, 2024

weilinquan commented Jan 4, 2024 •

edited

Loading