flux fp4 example(WIP)

[lanluo-nvidia]

Description

Please include a summary of the change and which issue is fixed. Please also include relevant motivation and context. List any dependencies that are required for this change.

Fixes # (issue)

Type of change

Please delete options that are not relevant and/or add your own.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

Checklist:

• My code follows the style guidelines of this project (You can use the linters)
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas and hacks
• I have made corresponding changes to the documentation
• I have added tests to verify my fix or my feature
• New and existing unit tests pass locally with my changes
• I have added the relevant labels to my PR in so that relevant reviewers are notified

[github-actions]

There are some changes that do not conform to Python style guidelines:

--- /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/addmm.py	2025-05-28 16:10:39.268834+00:00
+++ /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/addmm.py	2025-05-28 16:11:01.327800+00:00
@@ -6,10 +6,11 @@
from torch_tensorrt.dynamo._SourceIR import SourceIR
from torch_tensorrt.dynamo.conversion import impl
from torch_tensorrt.dynamo.conversion._ConversionContext import ConversionContext
from torch_tensorrt.fx.types import TRTTensor
import os
+

def addmm(
    ctx: ConversionContext,
    target: Target,
    source_ir: Optional[SourceIR],
--- /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/_TRTInterpreter.py	2025-05-28 16:10:39.267834+00:00
+++ /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/_TRTInterpreter.py	2025-05-28 16:11:01.908540+00:00
@@ -272,17 +272,23 @@
            builder_config.set_memory_pool_limit(
                trt.MemoryPoolType.DLA_GLOBAL_DRAM,
                self.compilation_settings.dla_global_dram_size,
            )

-        if not self.compilation_settings.use_explicit_typing and dtype.float16 in self.compilation_settings.enabled_precisions:
+        if (
+            not self.compilation_settings.use_explicit_typing
+            and dtype.float16 in self.compilation_settings.enabled_precisions
+        ):
            builder_config.set_flag(trt.BuilderFlag.FP16)

        if dtype.int8 in self.compilation_settings.enabled_precisions:
            builder_config.set_flag(trt.BuilderFlag.INT8)

-        if not self.compilation_settings.use_explicit_typing and dtype.fp8 in self.compilation_settings.enabled_precisions:
+        if (
+            not self.compilation_settings.use_explicit_typing
+            and dtype.fp8 in self.compilation_settings.enabled_precisions
+        ):
            builder_config.set_flag(trt.BuilderFlag.FP8)

        if dtype.bfloat16 in self.compilation_settings.enabled_precisions:
            builder_config.set_flag(trt.BuilderFlag.BF16)

--- /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/permutation.py	2025-05-28 16:10:39.269834+00:00
+++ /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/permutation.py	2025-05-28 16:11:01.949586+00:00
@@ -13,10 +13,11 @@
)
from torch_tensorrt.dynamo.conversion.impl.shape import get_shape_with_dynamic_shape
from torch_tensorrt.fx.types import TRTTensor
import os

+
def permute(
    ctx: ConversionContext,
    target: Target,
    source_ir: Optional[SourceIR],
    name: str,

[github-actions]

There are some changes that do not conform to Python style guidelines:

--- /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/quantize.py	2025-06-08 16:14:53.013799+00:00
+++ /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/quantize.py	2025-06-08 16:15:15.490934+00:00
@@ -94,6 +94,6 @@
        if axis is not None:
            dequantize_layer.axis = axis
        set_layer_name(dequantize_layer, target, name + "_dequantize", source_ir)
        dq_output = dequantize_layer.get_output(0)

-        return dq_output
\ No newline at end of file
+        return dq_output

[github-actions]

There are some changes that do not conform to Python style guidelines:

--- /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/quantize.py	2025-06-09 16:52:54.851163+00:00
+++ /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/quantize.py	2025-06-09 16:53:17.995276+00:00
@@ -94,6 +94,6 @@
        if axis is not None:
            dequantize_layer.axis = axis
        set_layer_name(dequantize_layer, target, name + "_dequantize", source_ir)
        dq_output = dequantize_layer.get_output(0)

-        return dq_output
\ No newline at end of file
+        return dq_output

[github-actions]

There are some changes that do not conform to Python style guidelines:

--- /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/quantize.py	2025-06-09 23:54:44.134169+00:00
+++ /home/runner/work/TensorRT/TensorRT/py/torch_tensorrt/dynamo/conversion/impl/quantize.py	2025-06-09 23:55:10.573434+00:00
@@ -94,6 +94,6 @@
        if axis is not None:
            dequantize_layer.axis = axis
        set_layer_name(dequantize_layer, target, name + "_dequantize", source_ir)
        dq_output = dequantize_layer.get_output(0)

-        return dq_output
\ No newline at end of file
+        return dq_output