Adding installation and usage to readme

Summary:

Test Plan:

Reviewers:

Subscribers:

Tasks:

Tags:

ghstack-source-id: f63aa75
Pull Request resolved: #7

Author: HDCharles

README.md

@@ -4,11 +4,52 @@ The torchao package contains apis and workflows used to apply AO techniques like
 
 ## Installation
 
-tbd
+clone repository and install package:
+
+```
+git clone https://github.com/pytorch-labs/ao
+cd ao
+python setup.py install
+```
+
+verify installation:
+
+```
+pip list | grep torchao
+```
+
+should show
+```
+torchao                            0.0.1                   <install dir>
+```
 
 ## Usage
 
-tbd
+Relevant APIs can be found in torchao.quantization.quant_api
+
+Example
+
+```
+import torch
+from torchao.quantization import quant_api
+
+# some user model
+model = torch.nn.Sequential(torch.nn.Linear(32, 64)).cuda().to(torch.bfloat16)
+# some example input
+input = torch.randn(32,32, dtype=torch.bfloat16, device='cuda')
+
+# convert linear modules to quantized linear modules
+# insert quantization method/api of choice
+quant_api.apply_weight_only_int8_quant(model)
+# quant_api.apply_dynamic_quant(model)
+# quant_api.change_linear_weights_to_dqtensors(model)
+
+# compile the model to improve performance
+torch.compile(model, mode='max-autotune')
+model(input)
+```
+
+### A16W8 WeightOnly Quantization
 
 ## License