Update README.md: add supported models

README.md

@@ -15,3 +15,47 @@ in addition, tachikoma's TVM fork is useful for:
 - converting a floating-point neural network or a framework-prequantized model into an integer-only form
 - generating a computational trace binary respecting the tachikoma standard.
 - as a proof of concept, how tachikoma can be used in ZKP systems. We will be implementing a simple graph runtime on top of the tachikoma standard, as well as a circuit builder in ZEXE. The code will be available here: https://github.com/zk-ml/tachikoma-poc-runtime
+
+---------------
+
+## MRT + ZKML
+
+*Developed by CortexLabs Team. All models are derived from the gluoncv model zoo.*
+
+Current MRT has supported most CNN classification models and limited detection models. More details are listed as following:
+
+### Classification Model
+
+*The main test code is located at tests/models/classification/test.main.py.*
+
+- "resnet18"                 # passed
+
+  Iteration: 19 | from_expr: Top1/5: 77.50%,94.69% | sim: Top1/5: 77.50%,94.69% | clip: Top1/5: 77.81%,94.69% | round: Top1/5: 75.62%,94.06% | quantized: Top1/5: 75.31%,94.06% |
+
+- "mobilenet_v2"             # passed
+- "alexnet"                  # passed
+
+  Iteration: 19 | from_expr: Top1/5: 66.88%,88.44% | sim: Top1/5: 66.88%,88.44% | clip: Top1/5: 67.19%,88.44% | round: Top1/5: 66.56%,89.06% | quantized: Top1/5: 66.56%,89.06% |
+
+- "densenet121"              # passed
+
+  Iteration:  19 | from_expr: Top1/5: 83.75%,96.56% | sim: Top1/5: 83.75%,96.56% | clip: Top1/5: 83.75%,96.88% | round: Top1/5: 51.25%,81.25% | quantized: Top1/5: 50.31%,81.25% |
+
+- "squeezenet1_0"            # passed
+
+  Iteration:  19 | from_expr: Top1/5: 70.31%,91.56% | sim: Top1/5: 70.31%,91.56% | clip: Top1/5: 70.31%,91.56% | round: Top1/5: 59.06%,85.94% | quantized: Top1/5: 59.38%,85.94% |
+
+- "vgg11"                    # passed
+
+  Iteration:  19 | from_expr: Top1/5: 79.06%,95.00% | sim: Top1/5: 79.06%,95.00% | clip: Top1/5: 79.06%,95.00% | round: Top1/5: 77.50%,95.94% | quantized: Top1/5: 77.50%,95.94% |
+
+- "shufflenet_v2_x0_5"       # passed
+
+  Iteration:  19 | from_expr: Top1/5: 73.12%,90.00% | sim: Top1/5: 73.12%,90.00% | clip: Top1/5: 73.75%,89.69% | round: Top1/5: 0.94%,4.38% | quantized: Top1/5: 0.94%,4.38% |
+
+### Detection Models
+
+*The main test code is located at tests/models/detection/test.mxnet.py.*
+
+- "mxnet_ssd_512_resnet50_v1_voc"     # passed
+- "yolo3_darknet53_voc"               # passed