FP8 2D forward convolution using rocMLIR

[umangyadav]

FP8 convolutions that uses rocMLIR.

Uses rocMLIR from SHA : 5085343bca363109ae9ebabb7ca2b65c52bc861c

ROCm/rocMLIR#1336

regular convolution takes in both inputs as FP8 and generates FP8 output. Internally on hardware it will do accumulation in Fp32 but final result is converted back to fp8 using downcasting.

quant_convolution takes in both inputs as fp8 and generates FP32 output. This version can utilize QDQ quantization scheme to use scales to downcast FP32 output to fp8.

rocMLIR fp8 convolution are limited to 2d forward convolutions only. Therefore only have added tests for those.
3D convolutions, 1D convolution, backwards convolution (transposed convolutions) "tests" therefore are not enabled.

convert fusion with mlir-convolution is not compiling on non-MI300 hardware therefore disabled it.

For now i've kept both versions of convolutions (quant and regular). In future if there is no use for regular fp8 convolution then it can be removed.

Testing : make check passes on MI300.

depends on #2473

[migraphx-bot]

     ✅ bert-mrpc-onnx: PASSED: MIGraphX meets tolerance

     ✅ bert-mrpc-tf: PASSED: MIGraphX meets tolerance

     ✅ pytorch-examples-wlang-gru: PASSED: MIGraphX meets tolerance

     ✅ pytorch-examples-wlang-lstm: PASSED: MIGraphX meets tolerance

     ✅ torchvision-resnet50_1: PASSED: MIGraphX meets tolerance

     ✅ torchvision-inceptionv3_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-dpn92_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-resnext101_1: PASSED: MIGraphX meets tolerance

     ✅ slim-vgg16_1: PASSED: MIGraphX meets tolerance

     ✅ slim-mobilenet_1: PASSED: MIGraphX meets tolerance

     ✅ slim-inceptionv4_1: PASSED: MIGraphX meets tolerance

     ✅ dlrm-criteoterabyte: PASSED: MIGraphX meets tolerance

     ✅ agentmodel: PASSED: MIGraphX meets tolerance

     ✅ unet: PASSED: MIGraphX meets tolerance

     ✅ resnet50v1: PASSED: MIGraphX meets tolerance

🔴bert_base_cased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

     ✅ bert_large_uncased_fp16: PASSED: MIGraphX meets tolerance

     ✅ bert_large: PASSED: MIGraphX meets tolerance

🔴distilgpt2_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

[TedThemistokleous]

LGTM