opengradient 0.2.7__py3-none-any.whl → 0.3.0__py3-none-any.whl

opengradient/__init__.py

@@ -1,16 +1,16 @@
 from .client import Client
-from .exceptions import OpenGradientError, FileNotFoundError, UploadError, InferenceError, ResultRetrievalError
-from .types import ModelInput, InferenceMode, Number, NumberTensor, StringTensor, ModelOutput
-
-__version__ = "0.2.7"
+from .defaults import *
+from .types import InferenceMode
+from typing import List, Dict
+__version__ = "0.3.0"
 
 _client = None
 
-def init(private_key="cd09980ef6e280afc3900d2d6801f9e9c5d858a5deaeeab74a65643f5ff1a4c1",
-         rpc_url="http://18.218.115.248:8545",
-         contract_address="0x350E0A430b2B1563481833a99523Cfd17a530e4e",
-         email="test@test.com",
-         password="Test-123"):
+def init(email: str,
+         password: str,
+         private_key=DEFAULT_PRIVATE_KEY,
+         rpc_url=DEFAULT_RPC_URL,
+         contract_address=DEFAULT_INFERENCE_CONTRACT_ADDRESS):
     global _client
     _client = Client(private_key=private_key, rpc_url=rpc_url, contract_address=contract_address, email=email, password=password)
 
@@ -19,10 +19,18 @@ def upload(model_path, model_name, version):
         raise RuntimeError("OpenGradient client not initialized. Call og.init() first.")
     return _client.upload(model_path, model_name, version)
 
-def create_model(model_name, model_desc):
+def create_model(model_name: str, model_desc: str, model_path: str = None):
     if _client is None:
         raise RuntimeError("OpenGradient client not initialized. Call og.init() first.")
-    return _client.create_model(model_name, model_desc)
+    
+    result = _client.create_model(model_name, model_desc)
+    
+    if model_path:
+        version = "0.01"
+        upload_result = _client.upload(model_path, model_name, version)
+        result["upload"] = upload_result
+    
+    return result
 
 def create_version(model_name, notes=None, is_major=False):
     if _client is None:
@@ -37,4 +45,9 @@ def infer(model_cid, inference_mode, model_input):
 def login(email: str, password: str):
     if _client is None:
         raise RuntimeError("OpenGradient client not initialized. Call og.init() first.")
-    return _client.login(email, password)
+    return _client.login(email, password)
+
+def list_files(model_name: str, version: str) -> List[Dict]:
+    if _client is None:
+        raise RuntimeError("OpenGradient client not initialized. Call og.init() first.")
+    return _client.list_files(model_name, version)

opengradient/abi/inference.abi

@@ -1 +1 @@
-[{"anonymous":false,"inputs":[{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct Number[]","name":"values","type":"tuple[]"}],"internalType":"struct NumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct StringTensor[]","name":"strings","type":"tuple[]"},{"internalType":"bool","name":"is_simulation_result","type":"bool"}],"indexed":false,"internalType":"struct ModelOutput","name":"output","type":"tuple"}],"name":"InferenceResult","type":"event"},{"inputs":[{"internalType":"string","name":"modelId","type":"string"},{"internalType":"enum IInference.ModelInferenceMode","name":"inferenceMode","type":"uint8"},{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct Number[]","name":"values","type":"tuple[]"}],"internalType":"struct NumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct StringTensor[]","name":"strings","type":"tuple[]"}],"internalType":"struct ModelInput","name":"modelInput","type":"tuple"}],"name":"run","outputs":[{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct Number[]","name":"values","type":"tuple[]"}],"internalType":"struct NumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct StringTensor[]","name":"strings","type":"tuple[]"},{"internalType":"bool","name":"is_simulation_result","type":"bool"}],"internalType":"struct ModelOutput","name":"","type":"tuple"}],"stateMutability":"nonpayable","type":"function"}]
+[{"anonymous":false,"inputs":[{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct TensorLib.Number[]","name":"values","type":"tuple[]"},{"internalType":"uint32[]","name":"shape","type":"uint32[]"}],"internalType":"struct TensorLib.MultiDimensionalNumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct TensorLib.StringTensor[]","name":"strings","type":"tuple[]"},{"internalType":"bool","name":"is_simulation_result","type":"bool"}],"indexed":false,"internalType":"struct ModelOutput","name":"output","type":"tuple"}],"name":"InferenceResult","type":"event"},{"inputs":[{"internalType":"string","name":"modelId","type":"string"},{"internalType":"enum ModelInferenceMode","name":"inferenceMode","type":"uint8"},{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct TensorLib.Number[]","name":"values","type":"tuple[]"},{"internalType":"uint32[]","name":"shape","type":"uint32[]"}],"internalType":"struct TensorLib.MultiDimensionalNumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct TensorLib.StringTensor[]","name":"strings","type":"tuple[]"}],"internalType":"struct ModelInput","name":"modelInput","type":"tuple"}],"name":"run","outputs":[{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct TensorLib.Number[]","name":"values","type":"tuple[]"},{"internalType":"uint32[]","name":"shape","type":"uint32[]"}],"internalType":"struct TensorLib.MultiDimensionalNumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct TensorLib.StringTensor[]","name":"strings","type":"tuple[]"},{"internalType":"bool","name":"is_simulation_result","type":"bool"}],"internalType":"struct ModelOutput","name":"","type":"tuple"}],"stateMutability":"nonpayable","type":"function"}]

opengradient/abi/llm.abi

@@ -0,0 +1 @@
+[{"anonymous":false,"inputs":[{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct TensorLib.Number[]","name":"values","type":"tuple[]"},{"internalType":"uint32[]","name":"shape","type":"uint32[]"}],"internalType":"struct TensorLib.MultiDimensionalNumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct TensorLib.StringTensor[]","name":"strings","type":"tuple[]"},{"internalType":"bool","name":"is_simulation_result","type":"bool"}],"indexed":false,"internalType":"struct ModelOutput","name":"output","type":"tuple"}],"name":"InferenceResult","type":"event"},{"anonymous":false,"inputs":[{"components":[{"internalType":"string","name":"answer","type":"string"}],"indexed":false,"internalType":"struct LlmResponse","name":"response","type":"tuple"}],"name":"LLMResult","type":"event"},{"inputs":[{"internalType":"string","name":"modelId","type":"string"},{"internalType":"enum ModelInferenceMode","name":"inferenceMode","type":"uint8"},{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct TensorLib.Number[]","name":"values","type":"tuple[]"},{"internalType":"uint32[]","name":"shape","type":"uint32[]"}],"internalType":"struct TensorLib.MultiDimensionalNumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct TensorLib.StringTensor[]","name":"strings","type":"tuple[]"}],"internalType":"struct ModelInput","name":"modelInput","type":"tuple"}],"name":"run","outputs":[{"components":[{"components":[{"internalType":"string","name":"name","type":"string"},{"components":[{"internalType":"int128","name":"value","type":"int128"},{"internalType":"int128","name":"decimals","type":"int128"}],"internalType":"struct TensorLib.Number[]","name":"values","type":"tuple[]"},{"internalType":"uint32[]","name":"shape","type":"uint32[]"}],"internalType":"struct TensorLib.MultiDimensionalNumberTensor[]","name":"numbers","type":"tuple[]"},{"components":[{"internalType":"string","name":"name","type":"string"},{"internalType":"string[]","name":"values","type":"string[]"}],"internalType":"struct TensorLib.StringTensor[]","name":"strings","type":"tuple[]"},{"internalType":"bool","name":"is_simulation_result","type":"bool"}],"internalType":"struct ModelOutput","name":"","type":"tuple"}],"stateMutability":"nonpayable","type":"function"},{"inputs":[{"components":[{"internalType":"enum LlmInferenceMode","name":"mode","type":"uint8"},{"internalType":"string","name":"modelCID","type":"string"},{"internalType":"string","name":"prompt","type":"string"},{"internalType":"uint32","name":"max_tokens","type":"uint32"},{"internalType":"string[]","name":"stop_sequence","type":"string[]"},{"internalType":"uint32","name":"temperature","type":"uint32"}],"internalType":"struct LlmInferenceRequest","name":"request","type":"tuple"}],"name":"runLLM","outputs":[{"components":[{"internalType":"string","name":"answer","type":"string"}],"internalType":"struct LlmResponse","name":"","type":"tuple"}],"stateMutability":"nonpayable","type":"function"}]

opengradient/account.py

@@ -0,0 +1,37 @@
+from eth_account import Account
+import secrets
+from collections import namedtuple
+import os
+import hashlib
+
+EthAccount = namedtuple('EthAccount', ['address', 'private_key'])
+
+
+def generate_eth_account() -> EthAccount:
+    user_seed = _get_user_random_seed()
+    private_key = _generate_secure_private_key(user_seed)
+
+    # derive account
+    account = Account.from_key(private_key)
+
+    # get the public key (address)
+    public_key = account.address
+
+    return EthAccount(address=public_key, private_key=private_key)
+
+def _get_user_random_seed():
+    print("Please type a random string of characters (the longer and more random, the better):")
+    print("> ", end="")  # Add a '>' prompt on a new line
+    return input().encode()
+
+def _generate_secure_private_key(user_input):
+    # Combine multiple sources of entropy
+    system_random = secrets.token_bytes(32)
+    os_urandom = os.urandom(32)
+    timestamp = str(secrets.randbits(256)).encode()
+    
+    # Add user input to the entropy sources
+    combined = system_random + os_urandom + timestamp + user_input
+    
+    # Hash the combined entropy
+    return hashlib.sha256(combined).hexdigest()

opengradient/cli.py

@@ -1,18 +1,30 @@
 import click
-import os
 import opengradient
 import json
 import ast
 from pathlib import Path
+import logging
+from pprint import pformat
+import webbrowser
+import sys
+
 from .client import Client
-from opengradient.types import InferenceMode, ModelInput
+from .defaults import *
+from .types import InferenceMode
+from .account import EthAccount, generate_eth_account
+
+OG_CONFIG_FILE = Path.home() / '.opengradient_config.json'
+
 
-# Environment variable names
-API_KEY_ENV = 'OPENGRADIENT_API_KEY'
-RPC_URL_ENV = 'OPENGRADIENT_RPC_URL'
-CONTRACT_ADDRESS_ENV = 'OPENGRADIENT_CONTRACT_ADDRESS'
-EMAIL_ENV = 'OPENGRADIENT_EMAIL'
-PASSWORD_ENV = 'OPENGRADIENT_PASSWORD'
+def load_og_config():
+    if OG_CONFIG_FILE.exists():
+        with OG_CONFIG_FILE.open('r') as f:
+            return json.load(f)
+    return {}
+
+def save_og_config(ctx):
+    with OG_CONFIG_FILE.open('w') as f:
+        json.dump(ctx.obj, f)
 
 # Convert string to dictionary click parameter typing
 class DictParamType(click.ParamType):
@@ -39,122 +51,231 @@ Dict = DictParamType()
 
 # Support inference modes
 InferenceModes = {
-    "VANILLA": opengradient.InferenceMode.VANILLA,
-    "ZKML": opengradient.InferenceMode.ZKML,
-    "TEE": opengradient.InferenceMode.TEE,
+    "VANILLA": InferenceMode.VANILLA,
+    "ZKML": InferenceMode.ZKML,
+    "TEE": InferenceMode.TEE,
 }
 
-# TODO (Kyle): Once we're farther into development, we should remove the defaults for these options
+
+def initialize_config(ctx):
+    """Interactively initialize OpenGradient config"""
+    if ctx.obj:  # Check if config data already exists
+        click.echo("A config already exists. Please run 'opengradient config clear' first if you want to reinitialize.")
+        click.echo("You can view your current config with 'opengradient config show'.")
+
+    click.echo("Initializing OpenGradient config...")
+    click.secho(f"Config will be stored in: {OG_CONFIG_FILE}", fg='cyan')
+
+    # Check if user has an existing account
+    has_account = click.confirm("Do you already have an OpenGradient account?", default=True)
+
+    if not has_account:
+        eth_account = create_account_impl()
+        if eth_account is None:
+            click.echo("Account creation cancelled. Config initialization aborted.")
+            return
+        ctx.obj['private_key'] = eth_account.private_key
+    else:
+        ctx.obj['private_key'] = click.prompt("Enter your OpenGradient private key", type=str)
+
+    # Make email and password optional
+    email = click.prompt("Enter your OpenGradient Hub email address (optional, press Enter to skip)", 
+                         type=str, default='', show_default=False)
+    ctx.obj['email'] = email if email else None
+    password = click.prompt("Enter your OpenGradient Hub password (optional, press Enter to skip)", 
+                            type=str, hide_input=True, default='', show_default=False)
+    ctx.obj['password'] = password if password else None
+
+    ctx.obj['rpc_url'] = DEFAULT_RPC_URL
+    ctx.obj['contract_address'] = DEFAULT_INFERENCE_CONTRACT_ADDRESS
+    
+    save_og_config(ctx)
+    click.echo("Config has been saved.")
+    click.secho("You can run 'opengradient config show' to see configs.", fg='green')
+
+
 @click.group()
-@click.option('--api_key', 
-              envvar=API_KEY_ENV, 
-              help='Your OpenGradient private key', 
-              default="cd09980ef6e280afc3900d2d6801f9e9c5d858a5deaeeab74a65643f5ff1a4c1")
-@click.option('--rpc_url', 
-              envvar=RPC_URL_ENV, 
-              help='OpenGradient RPC URL address', 
-              default="http://18.218.115.248:8545")
-@click.option('--contract_address', 
-              envvar=CONTRACT_ADDRESS_ENV, 
-              help='OpenGradient inference contract address', 
-              default="0x350E0A430b2B1563481833a99523Cfd17a530e4e")
-@click.option('--email', 
-              envvar=EMAIL_ENV,
-              help='Your OpenGradient Hub email address -- not required for inference', 
-              default="test@test.com")
-@click.option('--password', 
-              envvar=PASSWORD_ENV, 
-              help='Your OpenGradient Hub password -- not required for inference', 
-              default="Test-123")
 @click.pass_context
-def cli(ctx, api_key, rpc_url, contract_address, email, password):
-    """CLI for OpenGradient SDK"""
-    if not api_key:
-        click.echo("Please provide an API key via flag or setting environment variable OPENGRADIENT_API_KEY")
-    if not rpc_url:
-        click.echo("Please provide a RPC URL via flag or setting environment variable OPENGRADIENT_RPC_URL")
-    if not contract_address:
-        click.echo("Please provide a contract address via flag or setting environment variable OPENGRADIENT_CONTRACT_ADDRESS")
-    if not api_key or not rpc_url or not contract_address:
+def cli(ctx):
+    """CLI for OpenGradient SDK. Visit https://docs.opengradient.ai/developers/python_sdk/ for more documentation."""
+    # Load existing config 
+    ctx.obj = load_og_config()
+
+    no_client_commands = ['config', 'create-account', 'version']
+
+    # Only create client if this is not a config management command
+    if ctx.invoked_subcommand in no_client_commands:
+        return
+
+    if all(key in ctx.obj for key in ['private_key', 'rpc_url', 'contract_address']):
+        try:
+            ctx.obj['client'] = Client(private_key=ctx.obj['private_key'], 
+                                        rpc_url=ctx.obj['rpc_url'],
+                                        contract_address=ctx.obj['contract_address'],
+                                        email=ctx.obj.get('email'),
+                                        password=ctx.obj.get('password'))
+        except Exception as e:
+            click.echo(f"Failed to create OpenGradient client: {str(e)}")
+            ctx.exit(1)
+    else:
+        click.echo("Insufficient information to create client. Some commands may not be available.")
+        click.echo("Please run 'opengradient config clear' and/or 'opengradient config init' and to reinitialize your configs.")
         ctx.exit(1)
+
+
+@cli.group()
+def config():
+    """Manage your OpenGradient configuration (credentials etc)"""
+    pass
+
+
+@config.command()
+@click.pass_context
+def init(ctx):
+    """Initialize or reinitialize the OpenGradient config"""
+    initialize_config(ctx)
+
+
+@config.command()
+@click.pass_context
+def show(ctx):
+    """Display current config information"""
+    click.secho(f"Config file location: {OG_CONFIG_FILE}", fg='cyan')
+
+    if not ctx.obj:
+        click.echo("Config is empty. Run 'opengradient config init' to initialize it.")
         return
 
-    try:
-        ctx.obj = Client(private_key=api_key, 
-                         rpc_url=rpc_url,
-                         contract_address=contract_address,
-                         email=email,
-                         password=password)
-    except Exception as e:
-        click.echo(f"Failed to create OpenGradient client: {str(e)}")
+    click.echo("Current config:")
+    for key, value in ctx.obj.items():
+        if key != 'client':  # Don't display the client object
+            if key == 'password' and value is not None:
+                click.echo(f"{key}: {'*' * len(value)}")  # Mask the password
+            elif value is None:
+                click.echo(f"{key}: Not set")
+            else:
+                click.echo(f"{key}: {value}") 
 
-@cli.command()
+
+@config.command()
 @click.pass_context
-def client_settings(ctx):
-    """Display OpenGradient client settings"""
-    client = ctx.obj
-    if not client:
-        click.echo("Client not initialized")
-        ctx.exit(1)
-        
-    click.echo("Settings for OpenGradient client:")
-    click.echo(f"\tAPI key ({API_KEY_ENV}): {client.private_key}")
-    click.echo(f"\tRPC URL ({RPC_URL_ENV}): {client.rpc_url}")
-    click.echo(f"\tContract address ({CONTRACT_ADDRESS_ENV}): {client.contract_address}")
-    if client.user:
-        click.echo(f"\tEmail ({EMAIL_ENV}): {client.user["email"]}")
+def clear(ctx):
+    """Clear all saved configs"""
+    if not ctx.obj:
+        click.echo("No configs to clear.")
+        return
+
+    if click.confirm("Are you sure you want to clear all configs? This action cannot be undone.", abort=True):
+        ctx.obj.clear()
+        save_og_config(ctx)
+        click.echo("Configs cleared.")
     else:
-        click.echo(f"\tEmail: not set")
+        click.echo("Config clear cancelled.")
+
 
 @cli.command()
-@click.argument('model_path', type=Path)
-@click.argument('model_id', type=str)
-@click.argument('version_id', type=str)
+@click.option('--repo', '-r', '--name', 'repo_name', required=True, help='Name of the new model repository')
+@click.option('--description', '-d', required=True, help='Description of the model')
 @click.pass_obj
-def upload(client, model_path, model_id, version_id):
-    """Upload a model"""
+def create_model_repo(obj, repo_name: str, description: str):
+    """
+    Create a new model repository.
+
+    This command creates a new model repository with the specified name and description.
+    The repository name should be unique within your account.
+
+    Example usage:
+
+    \b
+    opengradient create-model-repo --name "my_new_model" --description "A new model for XYZ task"
+    opengradient create-model-repo -n "my_new_model" -d "A new model for XYZ task"
+    """
+    client: Client = obj['client']
+
     try:
-        result = client.upload(model_path, model_id, version_id)
-        click.echo(f"Model uploaded successfully: {result}")
+        result = client.create_model(repo_name, description)
+        click.echo(f"Model repository created successfully: {result}")
     except Exception as e:
-        click.echo(f"Error uploading model: {str(e)}")
+        click.echo(f"Error creating model: {str(e)}")
+
 
 @cli.command()
-@click.argument('model_name', type=str)
-@click.argument('model_desc', type=str)
+@click.option('--repo', '-r', 'repo_name', required=True, help='Name of the existing model repository')
+@click.option('--notes', '-n', help='Version notes (optional)')
+@click.option('--major', '-m', is_flag=True, default=False, help='Flag to indicate a major version update')
 @click.pass_obj
-def create_model(client, model_name, model_desc):
-    """Create a new model"""
+def create_version(obj, repo_name: str, notes: str, major: bool):
+    """Create a new version in an existing model repository.
+
+    This command creates a new version for the specified model repository. 
+    You can optionally provide version notes and indicate if it's a major version update.
+
+    Example usage:
+
+    \b
+    opengradient create-version --repo my_model_repo --notes "Added new feature X" --major
+    opengradient create-version -r my_model_repo -n "Bug fixes"
+    """
+    client: Client = obj['client']
+
     try:
-        result = client.create_model(model_name, model_desc)
-        click.echo(f"Model created successfully: {result}")
+        result = client.create_version(repo_name, notes, major)
+        click.echo(f"New version created successfully: {result}")
     except Exception as e:
-        click.echo(f"Error creating model: {str(e)}")
+        click.echo(f"Error creating version: {str(e)}")
+
 
 @cli.command()
-@click.argument('model_id', type=str)
-@click.option('--notes', type=str, default=None, help='Version notes')
-@click.option('--is-major', default=False, is_flag=True, help='Is this a major version')
+@click.argument('file_path', type=click.Path(exists=True, file_okay=True, dir_okay=False, readable=True, path_type=Path),
+                metavar='FILE_PATH')
+@click.option('--repo', '-r', 'repo_name', required=True, help='Name of the model repository')
+@click.option('--version', '-v', required=True, help='Version of the model (e.g., "0.01")')
 @click.pass_obj
-def create_version(client, model_id, notes, is_major):
-    """Create a new version of a model"""
+def upload_file(obj, file_path: Path, repo_name: str, version: str):
+    """
+    Upload a file to an existing model repository and version.
+
+    FILE_PATH: Path to the file you want to upload (e.g., model.onnx)
+
+    Example usage:
+
+    \b
+    opengradient upload-file path/to/model.onnx --repo my_model_repo --version 0.01
+    opengradient upload-file path/to/model.onnx -r my_model_repo -v 0.01
+    """
+    client: Client = obj['client']
+
     try:
-        result = client.create_version(model_id, notes, is_major)
-        click.echo(f"Version created successfully: {result}")
+        result = client.upload(file_path, repo_name, version)
+        click.echo(f"File uploaded successfully: {result}")
     except Exception as e:
-        click.echo(f"Error creating version: {str(e)}")
+        click.echo(f"Error uploading model: {str(e)}")
+
 
 @cli.command()
-@click.argument('model_cid', type=str)
-@click.argument('inference_mode', type=click.Choice(InferenceModes.keys()), default="VANILLA")
-@click.argument('input_data', type=Dict, required=False)
-@click.option('--input_file',
+@click.option('--model', '-m', 'model_cid', required=True, help='CID of the model to run inference on')
+@click.option('--mode', 'inference_mode', type=click.Choice(InferenceModes.keys()), default="VANILLA", 
+              help='Inference mode (default: VANILLA)')
+@click.option('--input', '-d', 'input_data', type=Dict, help='Input data for inference as a JSON string')
+@click.option('--input-file', '-f', 
               type=click.Path(exists=True, file_okay=True, dir_okay=False, readable=True, path_type=Path),
-              help="Optional file input for model inference -- must be JSON") 
+              help="JSON file containing input data for inference")
 @click.pass_context
-def infer(ctx, model_cid, inference_mode, input_data, input_file):
-    """Run inference on a model"""
-    client = ctx.obj
+def infer(ctx, model_cid: str, inference_mode: str, input_data, input_file: Path):
+    """
+    Run inference on a model.
+
+    This command runs inference on the specified model using the provided input data.
+    You must provide either --input or --input-file, but not both.
+
+    Example usage:
+
+    \b
+    opengradient infer --model Qm... --mode VANILLA --input '{"key": "value"}'
+    opengradient infer -m Qm... -i ZKML -f input_data.json
+    """
+    client: Client = ctx.obj['client']
+
     try:
         if not input_data and not input_file:
             click.echo("Must specify either input_data or input_file")
@@ -174,11 +295,12 @@ def infer(ctx, model_cid, inference_mode, input_data, input_file):
                 model_input = json.load(file)
             
         # Parse input data from string to dict
-        click.echo(f"Running {inference_mode} inference for {model_cid}...")
+        click.echo(f"Running {inference_mode} inference for model \"{model_cid}\"\n")
         tx_hash, model_output = client.infer(model_cid=model_cid, inference_mode=InferenceModes[inference_mode], model_input=model_input)
+
         click.secho("Success!", fg="green")
-        click.echo(f"\nTransaction Hash: \n{tx_hash}")
-        click.echo(f"\nInference result: \n{model_output}")
+        click.echo(f"Transaction hash: {tx_hash}")
+        click.echo(f"Inference result:\n{pformat(model_output, indent=2, width=120)}")
     except json.JSONDecodeError as e:
         click.echo(f"Error decoding JSON: {e}", err=True)
         click.echo(f"Error occurred on line {e.lineno}, column {e.colno}", err=True)
@@ -186,9 +308,83 @@ def infer(ctx, model_cid, inference_mode, input_data, input_file):
         click.echo(f"Error running inference: {str(e)}")
 
 
+@cli.command()
+def create_account():
+    """Create a new test account for OpenGradient inference and model management"""
+    create_account_impl()
+
+
+def create_account_impl() -> EthAccount:
+    click.echo("\n" + "=" * 50)
+    click.echo("OpenGradient Account Creation Wizard".center(50))
+    click.echo("=" * 50 + "\n")
+
+    click.echo("\n" + "-" * 50)
+    click.echo("Step 1: Create Account on OpenGradient Hub")
+    click.echo("-" * 50)
+
+    click.echo(f"Please create an account on the OpenGradient Hub")
+    webbrowser.open(DEFAULT_HUB_SIGNUP_URL, new=2)
+    click.confirm("Have you successfully created your account on the OpenGradient Hub?", abort=True)
+
+    click.echo("\n" + "-" * 50)
+    click.echo("Step 2: Generate Ethereum Account")
+    click.echo("-" * 50)
+    eth_account = generate_eth_account()
+    click.echo(f"Generated OpenGradient chain account with address: {eth_account.address}")
+
+    click.echo("\n" + "-" * 50)
+    click.echo("Step 3: Fund Your Account")
+    click.echo("-" * 50)
+    click.echo(f"Please fund your account clicking 'Request' on the Faucet website")
+    webbrowser.open(DEFAULT_OG_FAUCET_URL + eth_account.address, new=2)
+    click.confirm("Have you successfully funded your account using the Faucet?", abort=True)
+
+    click.echo("\n" + "=" * 50)
+    click.echo("Account Creation Complete!".center(50))
+    click.echo("=" * 50)
+    click.echo("\nYour OpenGradient account has been successfully created and funded.")
+    click.secho(f"Address: {eth_account.address}", fg='green')
+    click.secho(f"Private Key: {eth_account.private_key}", fg='green')
+    click.secho("\nPlease save this information for your records.\n", fg='cyan')
+
+    return eth_account
+
+
 @cli.command()
 def version():
+    """Return version of OpenGradient CLI"""
     click.echo(f"OpenGradient CLI version: {opengradient.__version__}")
 
+
+@cli.command()
+@click.option('--repo', '-r', 'repo_name', required=True, help='Name of the model repository')
+@click.option('--version', '-v', required=True, help='Version of the model (e.g., "0.01")')
+@click.pass_obj
+def list_files(client: Client, repo_name: str, version: str):
+    """
+    List files for a specific version of a model repository.
+
+    This command lists all files associated with the specified model repository and version.
+
+    Example usage:
+
+    \b
+    opengradient list-files --repo my_model_repo --version 0.01
+    opengradient list-files -r my_model_repo -v 0.01
+    """
+    try:
+        files = client.list_files(repo_name, version)
+        if files:
+            click.echo(f"Files for {repo_name} version {version}:")
+            for file in files:
+                click.echo(f"  - {file['name']} (Size: {file['size']} bytes)")
+        else:
+            click.echo(f"No files found for {repo_name} version {version}")
+    except Exception as e:
+        click.echo(f"Error listing files: {str(e)}")
+
+
 if __name__ == '__main__':
+    logging.getLogger().setLevel(logging.WARN)
     cli()

opengradient/client.py

@@ -11,8 +11,6 @@ from typing import Dict, Tuple, Union, List
 from web3.exceptions import ContractLogicError
 import firebase
 
-logging.basicConfig(level=logging.INFO)
-
 class Client:
     FIREBASE_CONFIG = {
         "apiKey": "AIzaSyDUVckVtfl-hiteBzPopy1pDD8Uvfncs7w",
@@ -23,7 +21,7 @@ class Client:
         "databaseURL": ""
     }
     
-    def __init__(self, private_key: str, rpc_url: str, contract_address: str, email: str = "test@test.com", password: str = "Test-123"):
+    def __init__(self, private_key: str, rpc_url: str, contract_address: str, email: str, password: str):
         """
         Initialize the Client with private key, RPC URL, and contract address.
 
@@ -43,13 +41,22 @@ class Client:
         self.firebase_app = firebase.initialize_app(self.FIREBASE_CONFIG)
         self.auth = self.firebase_app.auth()
         self.user = None
+        self.email = email
+        self.password = password
+        
+        logging.debug("Initialized client with parameters:\n"
+                      "private key: %s\n"
+                      "RPC URL: %s\n"
+                      "Contract Address: %s\n",
+                      private_key, rpc_url, contract_address)
 
         abi_path = os.path.join(os.path.dirname(__file__), 'abi', 'inference.abi')
         with open(abi_path, 'r') as abi_file:
             inference_abi = json.load(abi_file)
         self.abi = inference_abi
 
-        self.login(email, password)
+        if email is not None:
+            self.login(email, password)
 
     def _initialize_web3(self):
         """
@@ -273,7 +280,6 @@ class Client:
             logging.error(f"Request exception during upload: {str(e)}")
             if hasattr(e, 'response') and e.response is not None:
                 logging.error(f"Response status code: {e.response.status_code}")
-                logging.error(f"Response headers: {e.response.headers}")
                 logging.error(f"Response content: {e.response.text[:1000]}...")  # Log first 1000 characters
             raise OpenGradientError(f"Upload failed due to request exception: {str(e)}", 
                                     status_code=e.response.status_code if hasattr(e, 'response') else None)
@@ -402,4 +408,49 @@ class Client:
             return self.user
         except Exception as e:
             logging.error(f"Authentication failed: {str(e)}")
-            raise
+            raise
+
+    def list_files(self, model_name: str, version: str) -> List[Dict]:
+        """
+        List files for a specific version of a model.
+
+        Args:
+            model_name (str): The unique identifier for the model.
+            version (str): The version identifier for the model.
+
+        Returns:
+            List[Dict]: A list of dictionaries containing file information.
+
+        Raises:
+            OpenGradientError: If the file listing fails.
+        """
+        if not self.user:
+            raise ValueError("User not authenticated")
+
+        url = f"https://api.opengradient.ai/api/v0/models/{model_name}/versions/{version}/files"
+        headers = {
+            'Authorization': f'Bearer {self.user["idToken"]}'
+        }
+
+        logging.debug(f"List Files URL: {url}")
+        logging.debug(f"Headers: {headers}")
+
+        try:
+            response = requests.get(url, headers=headers)
+            response.raise_for_status()
+
+            json_response = response.json()
+            logging.info(f"File listing successful. Number of files: {len(json_response)}")
+            
+            return json_response
+
+        except requests.RequestException as e:
+            logging.error(f"File listing failed: {str(e)}")
+            if hasattr(e, 'response') and e.response is not None:
+                logging.error(f"Response status code: {e.response.status_code}")
+                logging.error(f"Response content: {e.response.text[:1000]}...")  # Log first 1000 characters
+            raise OpenGradientError(f"File listing failed: {str(e)}", 
+                                    status_code=e.response.status_code if hasattr(e, 'response') else None)
+        except Exception as e:
+            logging.error(f"Unexpected error during file listing: {str(e)}", exc_info=True)
+            raise OpenGradientError(f"Unexpected error during file listing: {str(e)}")

opengradient/defaults.py

@@ -0,0 +1,7 @@
+
+# Default variables
+DEFAULT_PRIVATE_KEY="cd09980ef6e280afc3900d2d6801f9e9c5d858a5deaeeab74a65643f5ff1a4c1"
+DEFAULT_RPC_URL="http://18.218.115.248:8545"
+DEFAULT_OG_FAUCET_URL="http://18.218.115.248:8080/?address="
+DEFAULT_HUB_SIGNUP_URL="https://hub.opengradient.ai/signup"
+DEFAULT_INFERENCE_CONTRACT_ADDRESS="0x75D0266DAb643417e9FFD828A1A31C1E039a966c"

opengradient/utils.py

@@ -40,11 +40,10 @@ def convert_to_model_input(inputs: Dict[str, np.ndarray]) -> Tuple[List[Tuple[st
     """
     Expect SDK input to be a dict with the format
         key: tensor name
-        value: np.array
+        value: np.array / list
 
-        Note: np.array types must be float or string. Ints currently not supported.
-
-    Return a tuple of (number tensors, string tensors) depending on the input type
+    Return a tuple of (number tensors, string tensors) depending on the input type. Each number and string tensor converted
+    to a numpy array and flattened and the shape saved.
     """
     logging.debug("Converting the following input dictionary to ModelInput: %s", inputs)
     number_tensors = []
@@ -59,19 +58,37 @@ def convert_to_model_input(inputs: Dict[str, np.ndarray]) -> Tuple[List[Tuple[st
             logging.debug(f"\tConverting single entry {tensor_data} to a list")
             tensor_data = np.array([tensor_data])
 
+        # Check if type is np array
+        if not isinstance(tensor_data, np.ndarray):
+            raise TypeError("Inference input must be list, numpy array, or type (str, int, float): %s" % type(tensor_data))
+
+        # Flatten list and retain shape
+        shape = tensor_data.shape
+        flat_data = tensor_data.flatten()
+        logging.debug("Shape and flattened data: %s, %s", shape, flat_data)
+
         # Parse into number and string tensors
         if issubclass(tensor_data.dtype.type, np.floating):
-            input = (tensor_name, [convert_to_fixed_point(i) for i in tensor_data])
+            # Convert to fixed-point tuples
+            data_type = np.dtype([('value', int), ('decimal', int)])
+            converted_tensor_data = np.array([convert_to_fixed_point(i) for i in flat_data], dtype=data_type)
+            
+            input = (tensor_name, converted_tensor_data.tolist(), shape)
             logging.debug("\tFloating tensor input: %s", input)
 
             number_tensors.append(input)
         elif issubclass(tensor_data.dtype.type, np.integer):
-            input = (tensor_name, [convert_to_fixed_point(int(i)) for i in tensor_data])
+            # Convert to fixed-point tuples
+            data_type = np.dtype([('value', int), ('decimal', int)])
+            converted_tensor_data = np.array([convert_to_fixed_point(int(i)) for i in flat_data], dtype=data_type)
+
+            input = (tensor_name, converted_tensor_data.tolist(), shape)
             logging.debug("\tInteger tensor input: %s", input)
 
             number_tensors.append(input)
         elif issubclass(tensor_data.dtype.type, np.str_):
-            input = (tensor_name, [s for s in tensor_data])
+            # TODO (Kyle): Add shape into here as well
+            input = (tensor_name, [s for s in flat_data])
             logging.debug("\tString tensor input: %s", input)
 
             string_tensors.append(input)
@@ -85,6 +102,15 @@ def convert_to_model_input(inputs: Dict[str, np.ndarray]) -> Tuple[List[Tuple[st
     return number_tensors, string_tensors
 
 def convert_to_model_output(event_data: AttributeDict) -> Dict[str, np.ndarray]:
+    """
+    Converts inference output into a user-readable output. 
+    Expects the inference node to return a dict with the format:
+        key: output_name (str)
+        value: (output_array (list), shape (list)) (tuple)
+
+    We need to reshape each output array using the shape parameter in order to get the array
+    back into its original shape.
+    """
     logging.debug(f"Parsing event data: {event_data}")
         
     output_dict = {}
@@ -98,6 +124,7 @@ def convert_to_model_output(event_data: AttributeDict) -> Dict[str, np.ndarray]:
             logging.debug(f"Processing number tensor: {tensor}")
             if isinstance(tensor, AttributeDict):
                 name = tensor.get('name')
+                shape = tensor.get('shape')
                 values = []
                 # Convert from fixed point back into np.float32
                 for v in tensor.get('values', []):
@@ -105,7 +132,7 @@ def convert_to_model_output(event_data: AttributeDict) -> Dict[str, np.ndarray]:
                         values.append(convert_to_float32(value=int(v.get('value')), decimals=int(v.get('decimals'))))
                     else:
                         logging.warning(f"Unexpected number type: {type(v)}")
-                output_dict[name] = np.array(values)
+                output_dict[name] = np.array(values).reshape(shape)
             else:
                 logging.warning(f"Unexpected tensor type: {type(tensor)}")
 
@@ -114,8 +141,9 @@ def convert_to_model_output(event_data: AttributeDict) -> Dict[str, np.ndarray]:
             logging.debug(f"Processing string tensor: {tensor}")
             if isinstance(tensor, AttributeDict):
                 name = tensor.get('name')
+                shape = tensor.get('shape')
                 values = tensor.get('values', [])
-                output_dict[name] = values
+                output_dict[name] = np.array(values).reshape(shape)
             else:
                     logging.warning(f"Unexpected tensor type: {type(tensor)}")
     else:

{opengradient-0.2.7.dist-info → opengradient-0.3.0.dist-info}/METADATA

@@ -1,7 +1,7 @@
 Metadata-Version: 2.1
 Name: opengradient
-Version: 0.2.7
-Summary: A Python SDK for OpenGradient inference services
+Version: 0.3.0
+Summary: Python SDK for OpenGradient decentralized model management & inference services
 Author-email: OpenGradient <oliver@opengradient.ai>
 License: MIT License
         
@@ -135,47 +135,86 @@ Requires-Dist: websockets==13.1
 Requires-Dist: xattr==1.1.0
 Requires-Dist: yarl==1.13.1
 
-# OpenGradient Python SDK
+# OpenGradient SDK
 
-Python SDK for OpenGradient inference services.
+Python SDK for OpenGradient decentralized model management & inference services.
 
 ## Installation
-```
+```python
 pip install opengradient
 ```
 
 ## Quick Start
-```
+```python
 import opengradient as og
-og.init(private_key="x", rpc_url="y", contract_address="z")
-```
-
-### Sign in with Email
-```
-og.login(email="you@opengradient.ai", password="xyz")
+og.init(email="<email>", password="<password>")
 ```
 
 ### Create a Model
+```python
+og.create_model(model_name="<model_name>", model_desc="<model_description>")
 ```
-og.create_model(model_name="test-network-model", model_desc="testing upload to sdk")
+
+### Create a Model (with file upload)
+```python
+og.create_model(model_name="<model_name>", model_desc="<model_description>", model_path="<model_path>")
 ```
 
 ### Create a Version of a Model
-```
-og.create_version(model_name="test-network-model", notes="test notes")
+```python
+og.create_version(model_name="<model_name>", notes="<model_notes>")
 ```
 
 ### Upload Files to a Model
+```python
+og.upload(model_path="<model_path>", model_name="<model_name>", version="<version>")
 ```
-og.upload(model_path="local_path_to_your_model.onnx", model_name="test-network-model", version="0.01")
+
+### List Files of a Model Version
+```python
+og.list_files(model_name="<model_name>", version="<version>")
 ```
 
 ### Run Inference
-```
+```python
 inference_mode = og.InferenceMode.VANILLA
-inference_cid = og.infer(model_cid, model_inputs, inference_mode)
+og.infer(model_cid, model_inputs, inference_mode)
 ```
 
+## Using the CLI
+
+```bash
+export OPENGRADIENT_EMAIL="<email>"
+export OPENGRADIENT_PASSWORD="<password>"
 ```
-og.infer(model_id, inference_mode, model_input)
+
+#### Creating a Model
+```bash
+opengradient create_model "<model_name>" "<description>" 
+```
+- creating a model automatically initializes version `v0.01`
+
+#### Creating a Version
+```bash
+opengradient create_model "<model_name>" "<description>" 
+```
+
+#### Upload a File
+```bash
+opengradient upload "<model_path>" "<model_name>" "<version>" 
+```
+
+#### List Files of a Model Version
+```bash
+opengradient list_files "<model_name>" "<version>"
+```
+
+####  CLI infer using string 
+```bash
+opengradient infer QmbUqS93oc4JTLMHwpVxsE39mhNxy6hpf6Py3r9oANr8aZ VANILLA '{"num_input1":[1.0, 2.0, 3.0], "num_input2":10, "str_input1":["hello", "ONNX"], "str_input2":" world"}'
+```
+
+#### CLI infer using file path input
+```bash
+opengradient infer QmbUqS93oc4JTLMHwpVxsE39mhNxy6hpf6Py3r9oANr8aZ VANILLA --input_file input.json
 ```

opengradient-0.3.0.dist-info/RECORD

@@ -0,0 +1,16 @@
+opengradient/__init__.py,sha256=JtRa1lL5qn5slM2egGm76gVuft-qEHc9gtU6tbOPRG8,2028
+opengradient/account.py,sha256=s1C4hAtc8vcHObWjwxwlYJA041S6DTbr7-rK6qiWPsQ,1149
+opengradient/cli.py,sha256=YKctHMZhT_Y1fANWDnGo68QpIVLXqz5ifH5kQXIxD8A,14412
+opengradient/client.py,sha256=__KtU-i6EQqZNj8xX-yKoqb_V8Cdzgr9zHBY0m02Xfw,20141
+opengradient/defaults.py,sha256=5JNcTNfOmoP3DymEvBqZPQHUYCIslvxvLRtJLon0MkM,354
+opengradient/exceptions.py,sha256=v4VmUGTvvtjhCZAhR24Ga42z3q-DzR1Y5zSqP_yn2Xk,3366
+opengradient/types.py,sha256=EoJN-DkQrJ2WTUv8OenlrlWJWFY2jPGTl-T8C_OVjp8,1849
+opengradient/utils.py,sha256=F1Nj-GMNFQFxCtbGgWQq1RP4TSurbpQxJV3yKeEo1b0,6482
+opengradient/abi/inference.abi,sha256=u8FsW0s1YeRjUb9eLS1k_qh_5f_cwOdr0bii-tAdxh0,2683
+opengradient/abi/llm.abi,sha256=zhiPFyBT09EI3QU5DVoKHo7e8T9PFcfIQ3RHDYetm4M,3609
+opengradient-0.3.0.dist-info/LICENSE,sha256=xEcvQ3AxZOtDkrqkys2Mm6Y9diEnaSeQRKvxi-JGnNA,1069
+opengradient-0.3.0.dist-info/METADATA,sha256=A6qo4CPFnnCw1N_PK5yfxKS9hCxFUbpvKEiBWma8byw,7168
+opengradient-0.3.0.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
+opengradient-0.3.0.dist-info/entry_points.txt,sha256=yUKTaJx8RXnybkob0J62wVBiCp_1agVbgw9uzsmaeJc,54
+opengradient-0.3.0.dist-info/top_level.txt,sha256=oC1zimVLa2Yi1LQz8c7x-0IQm92milb5ax8gHBHwDqU,13
+opengradient-0.3.0.dist-info/RECORD,,

opengradient-0.2.7.dist-info/RECORD

@@ -1,13 +0,0 @@
-opengradient/__init__.py,sha256=BRxrOg8w_gPDiYpBbDAeEy-XFidmGNU-GvWozFw3vbo,1783
-opengradient/cli.py,sha256=EVb_JNvfaXWEalKKtfRknyw7tXko5qqKkHPwTR-W3cQ,7468
-opengradient/client.py,sha256=XJ5AGB518Mwmfdk1xeHyR51B2KvGvFHEoRjWLsr1x7M,18052
-opengradient/exceptions.py,sha256=v4VmUGTvvtjhCZAhR24Ga42z3q-DzR1Y5zSqP_yn2Xk,3366
-opengradient/types.py,sha256=EoJN-DkQrJ2WTUv8OenlrlWJWFY2jPGTl-T8C_OVjp8,1849
-opengradient/utils.py,sha256=95i5RVn-32MRsn00M21io8QHLtmEAoRbgueMhDh0TVk,5079
-opengradient/abi/inference.abi,sha256=HH2SmCJ_D4O0I-CFsln0vFHd2PU-A-fxgCnUtHg0ZQg,2373
-opengradient-0.2.7.dist-info/LICENSE,sha256=xEcvQ3AxZOtDkrqkys2Mm6Y9diEnaSeQRKvxi-JGnNA,1069
-opengradient-0.2.7.dist-info/METADATA,sha256=AXjww8-Sq9V1hil5VFOYGdOV9A7Yd60ytTGOtu2u1QQ,6085
-opengradient-0.2.7.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
-opengradient-0.2.7.dist-info/entry_points.txt,sha256=yUKTaJx8RXnybkob0J62wVBiCp_1agVbgw9uzsmaeJc,54
-opengradient-0.2.7.dist-info/top_level.txt,sha256=oC1zimVLa2Yi1LQz8c7x-0IQm92milb5ax8gHBHwDqU,13
-opengradient-0.2.7.dist-info/RECORD,,