Add command docs

Author: lantiga

docs/commands.md

@@ -1,41 +1,143 @@
 # RedisAI Commands
 
-## AI.TENSORSET - Set a tensor
-Stores a tensor of defined type (FLOAT, DOUBLE, INT8, INT16, INT32, INT64, UINT8, UINT16) with N dimensions (dim) and shape given by shape1..shapeN
+## AI.TENSORSET - set a tensor
+> Stores a tensor of defined type  with shape given by shape1..shapeN
 
-```sh
-AI.TENSORSET tensor_key data_type shape1..shapeN [BLOB data | VALUES val1..valN]
+```sql
+AI.TENSORSET tensor_key data_type shape1 shape2 ... [BLOB data | VALUES val1 val2 ...]
 ```
 
-## AI.TENSORGET - Get a tensor
+* tensor_key - key for storing the tensor
+* data_type - numeric data type of tensor elements, one of FLOAT, DOUBLE, INT8, INT16, INT32, INT64, UINT8, UINT16
+* shape - shape of the tensor, i.e. how many elements for each axis
 
-```sh
+Optional args:
+    * BLOB data - provide tensor content as a binary buffer
+    * VALUES val1 val2 - provide tensor content as individual values
+
+> If no BLOB or VALUES are specified, the tensor is allocated but not initialized to any value.
+
+### Example
+> Set a 2x2 tensor at `foo`
+> 1 2
+> 3 4
+
+```sql
+AI.TENSORSET foo FLOAT 2 2 VALUES 1 2 3 4
+```
+
+## AI.TENSORGET - set a tensor
+```sql
 AI.TENSORGET tensor_key [BLOB | VALUES | META]
 ```
 
-## AI.MODELSET - Set a model
-Stores a model provided as a protobuf blob. Backend is TF or TORCH. The TF backend requires the name of input and output nodes to be specified in INPUTS and OUTPUTS.
-```sh
+* tensor_key - key for the tensor
+* BLOB - return tensor content as a binary buffer
+* VALUES - return tensor content as a list of values
+* META - only return tensor meta data (datat type and shape)
+
+### Example
+> Get binary data for tensor at `foo`. Meta data is also returned.
+
+```sql
+AI.TENSORGET foo BLOB
+```
+
+## AI.MODELSET - set a model
+```sql
 AI.MODELSET model_key backend device [INPUTS name1 name2 ... OUTPUTS name1 name2 ...] model_blob
 ```
 
-## AI.MODELRUN - Run a model
-```sh
+* model_key - key for storing the model
+* backend - the backend corresponding to the model being set. Allowed values: `TF`, `TORCH`.
+* device - device where the model is loaded and where the computation will run. Allowed values: `CPU`, `GPU`.
+* INPUTS name1 name2 ... - name of the nodes in the provided graph corresponding to inputs [`TF` backend only]
+* OUTPUTS name1 name2 ... - name of the nodes in the provided graph corresponding to outputs [`TF` backend only]
+* model_blob - binary buffer containing the model protobuf saved from a supported backend
+
+### Example
+
+```sql
+AI.MODELSET resnet18 TORCH GPU < foo.pt
+```
+
+```sql
+AI.MODELSET resnet18 TF CPU INPUTS in1 OUTPUTS linear4 < foo.pt
+```
+
+## AI.MODELRUN - run a model
+```sql
 AI.MODELRUN model_key INPUTS input_key1 ... OUTPUTS output_key1 ...
 ```
 
-## AI.SCRIPTSET - Set a script
-Stores a TorchScript script provided as text.
-```sh
-AI.SCRIPTSET script_key device script_text
+* model_key - key for the model
+* INPUTS input_key1 ... - keys for tensors to use as inputs
+* OUTPUTS output_key2 ... - keys for storing output tensors
+
+> The request is queued and evaded asynchronously from a separate thread. The client blocks until the computation finishes.
+
+> If needed, input tensors are copied to the device specified in `AI.MODELSET` before execution.
+
+### Example
+
+```sql
+AI.MODELRUN resnet18 INPUTS image12 OUTPUTS label12
+```
+
+
+## AI.SCRIPTSET - set a script
+```sql
+AI.SCRIPTSET script_key device script_source
+```
+
+* script_key - key for storing the script
+* device - the device where the script will execute
+* script_source - a string containing TorchScript source code
+
+### Example
+
+> Given addtwo.txt as:
+
+```python
+def addtwo(a, b):
+    return a + b
 ```
 
-## AI.SCRIPTGET - Get a script
-```sh
+```sql
+AI.SCRIPTSET addscript GPU < addtwo.txt
+```
+
+## AI.SCRIPTGET - get a script
+
+```sql
 AI.SCRIPTGET script_key
 ```
 
-## AI.SCRIPTRUN - Run a script
-```sh
+* script_key - key for the script
+
+### Example
+
+```sql
+AI.SCRIPTGET addscript
+```
+
+
+## AI.SCRIPTRUN - run a script
+
+```sql
 AI.SCRIPTRUN script_key fn_name INPUTS input_key1 ... OUTPUTS output_key1 ...
 ```
+
+* tensor_key - key for the script
+* fn_name - name of the function to execute
+* INPUTS input_key1 ... - keys for tensors to use as inputs
+* OUTPUTS output_key1 ... - keys for storing output tensors
+
+> If needed, input tensors are copied to the device specified in `AI.SCRIPTSET` before execution.
+
+
+### Example
+
+```sql
+AI.SCRIPTRUN addscript addtwo INPUTS a b OUTPUTS c
+```