Add Float8Tensor

Summary:
* Added Float8Tensor that's using fbgemm kernels and scaled_mm:
    * per row activation + per row weight linear calling torch._scaled_mm op (for compatibilty with SM 8.9)
    * per tensor activation + per tensor weight quant linear calling torch._scaled_mm op (for compatibilty with SM 8.9)
    * per row activation + per row weight bmm calling torch.ops.fbgemm.f8f8bf16_rowwise_batched kernel (only works for SM 9.0+) can use batched scaled mm from torch when it's supported: pytorch/pytorch#157950
* dynamic quantization kwargs is added to the Float8Tensor directly
* Added QuantizeTensorKwargs and QuantizeTensorToFloat8Kwargs to store key word args for Float8Tensor.to_float8
* Updated Float8DynamicActivationFloat8WeightConfig and Float8WeightOnlyConfig to use Float8Tensor

Test Plan:
python test/dtypes/test_affine_quantized_float.py
python test/quantization/quantize_/workflows/float8/test_float8_tensor.py

Reviewers:

Subscribers:

Tasks:

Tags:

stack-info: PR: #2463, branch: jerryzh168/stack/9

Author: jerryzh168

.github/workflows/1xH100_tests.yml

@@ -25,7 +25,7 @@ jobs:
         include:
           - name: H100
             runs-on: linux.aws.h100
-            torch-spec: '--pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu126'
+            torch-spec: '--pre torch torchvision torchaudio fbgemm-gpu-genai --index-url https://download.pytorch.org/whl/nightly/cu126'
             gpu-arch-type: "cuda"
             gpu-arch-version: "12.4"
     permissions:
@@ -46,8 +46,8 @@ jobs:
         pip install uv
         pip install ${{ matrix.torch-spec }}
         uv pip install -r dev-requirements.txt
-        uv pip install vllm
         pip install .
         pytest test/integration --verbose -s
         pytest test/dtypes/test_affine_quantized_float.py --verbose -s
+        pytest test/quantization/quantize_/workflows/float8/test_float8_tensor.py --verbose -s
         ./test/float8/test_everything_single_gpu.sh

.github/workflows/1xL4_tests.yml

@@ -25,7 +25,7 @@ jobs:
         include:
           - name: SM-89
             runs-on: linux.g6.4xlarge.experimental.nvidia.gpu
-            torch-spec: '--pre torch --index-url https://download.pytorch.org/whl/nightly/cu126'
+            torch-spec: '--pre torch fbgemm-gpu-genai --index-url https://download.pytorch.org/whl/nightly/cu126'
             gpu-arch-type: "cuda"
             gpu-arch-version: "12.6"
     permissions:
@@ -46,8 +46,8 @@ jobs:
         pip install uv
         pip install ${{ matrix.torch-spec }}
         uv pip install -r dev-requirements.txt
-        uv pip install vllm
         pip install .
         pytest test/integration --verbose -s
         pytest test/dtypes/test_affine_quantized_float.py --verbose -s
         ./test/float8/test_everything_single_gpu.sh
+        pytest test/quantization/quantize_/workflows/float8/test_float8_tensor.py --verbose -s