opengradient 0.3.13__py3-none-any.whl → 0.3.15__py3-none-any.whl

opengradient/__init__.py

@@ -5,7 +5,7 @@ from .defaults import DEFAULT_INFERENCE_CONTRACT_ADDRESS, DEFAULT_RPC_URL
 from .types import InferenceMode, LLM
 from . import llm
 
-__version__ = "0.3.13"
+__version__ = "0.3.15"
 
 _client = None
 

opengradient/llm/chat.py

@@ -10,6 +10,9 @@ from langchain.schema import (
     ChatResult,
     ChatGeneration,
 )
+from langchain_core.messages.tool import (
+    ToolMessage
+)
 from langchain_core.callbacks.manager import CallbackManagerForLLMRun
 from langchain_core.tools import BaseTool
 from langchain_core.messages import ToolCall
@@ -70,24 +73,36 @@ class OpenGradientChatModel(BaseChatModel):
         run_manager: Optional[CallbackManagerForLLMRun] = None,
         **kwargs: Any,
     ) -> ChatResult:
-        chat_messages = []
+
+        sdk_messages = []
         for message in messages:
             if isinstance(message, SystemMessage):
-                chat_messages.append({"role": "system", "content": message.content})
+                sdk_messages.append({"role": "system", "content": message.content})
             elif isinstance(message, HumanMessage):
-                chat_messages.append({"role": "user", "content": message.content})
+                sdk_messages.append({"role": "user", "content": message.content})
             elif isinstance(message, AIMessage):
-                chat_messages.append({"role": "assistant", "content": message.content})
+                sdk_messages.append({
+                    "role": "assistant", 
+                    "content": message.content, 
+                    "tool_calls": [{
+                        "id": call["id"],
+                        "name": call["name"],
+                        "arguments": json.dumps(call["args"])
+                    } for call in message.tool_calls]})
+            elif isinstance(message, ToolMessage):
+                sdk_messages.append({"role": "tool", "content": message.content, "tool_call_id": message.tool_call_id})
+            else:
+                raise ValueError(f"Unexpected message type: {message}")
 
-        tx_hash, finish_reason, chat_response = self.client.llm_chat(
+        _, finish_reason, chat_response = self.client.llm_chat(
             model_cid=self.model_cid,
-            messages=chat_messages,
+            messages=sdk_messages,
             stop_sequence=stop,
             max_tokens=self.max_tokens,
             tools=self.tools
         )
 
-        if "tool_calls" in chat_response:
+        if "tool_calls" in chat_response and chat_response["tool_calls"]:
             tool_calls = []
             for tool_call in chat_response["tool_calls"]:
                 tool_calls.append(
@@ -104,7 +119,7 @@ class OpenGradientChatModel(BaseChatModel):
             )
         else:
             message = AIMessage(content=chat_response["content"])
-        
+
         return ChatResult(
             generations=[ChatGeneration(
                 message=message,

{opengradient-0.3.13.dist-info → opengradient-0.3.15.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: opengradient
-Version: 0.3.13
+Version: 0.3.15
 Summary: Python SDK for OpenGradient decentralized model management & inference services
 Project-URL: Homepage, https://opengradient.ai
 Author-email: OpenGradient <oliver@opengradient.ai>
@@ -183,6 +183,39 @@ og.infer(model_cid, model_inputs, inference_mode)
 ```
  - inference mode can be `VANILLA`, `ZKML`, or `TEE`
 
+### LLM Inference
+#### LLM Completion
+```python
+tx_hash, response = og.llm_completion(
+    model_cid='meta-llama/Meta-Llama-3-8B-Instruct',
+    prompt="Translate the following English text to French: 'Hello, how are you?'",
+    max_tokens=50,
+    temperature=0.0
+)
+``` 
+
+#### LLM Chat
+```python
+# create messages history
+messages = [
+    {
+        "role": "system",
+        "content": "You are a helpful AI assistant.",
+        "name": "HAL"
+    },
+    {
+        "role": "user",
+        "content": "Hello! How are you doing? Can you repeat my name?",
+    }]
+
+# run LLM inference
+tx_hash, finish_reason, message = og.llm_chat(
+  model_cid=og.LLM.MISTRAL_7B_INSTRUCT_V3,
+  messages=messages
+)
+```
+
+
 
 ## Using the CLI
 
@@ -223,8 +256,18 @@ opengradient infer QmbUqS93oc4JTLMHwpVxsE39mhNxy6hpf6Py3r9oANr8aZ VANILLA --inpu
 ```
 
 #### Run LLM Inference
+We also have explicit support for using LLMs through the completion and chat commands in the CLI.
+
+For example, you can run a competion inference with Llama-3 using the following command:
+
+``` bash
+opengradient completion --model "meta-llama/Meta-Llama-3-8B-Instruct" --prompt "hello who are you?" --max-tokens 50
+```
+
+Or you can use files instead of text input in order to simplify your command:
+
 ```bash
-opengradient llm --model "meta-llama/Meta-Llama-3-8B-Instruct" --prompt "Translate to French: Hello, how are you?" --max-tokens 50 --temperature 0.7
+opengradient chat --model "mistralai/Mistral-7B-Instruct-v0.3" --messages-file messages.json --tools-file tools.json --max-tokens 200
 ```
 
 For more information read the OpenGradient [documentation](https://docs.opengradient.ai/).

{opengradient-0.3.13.dist-info → opengradient-0.3.15.dist-info}/RECORD

@@ -1,4 +1,4 @@
-opengradient/__init__.py,sha256=5aeZn5QVFepwTyjcO7GBHGlYUVwJCQa8QWYSHUaHqdM,4015
+opengradient/__init__.py,sha256=ATwimOCIZli1u5WKTMfhtH3VDeACqgK29BFAihyquvM,4015
 opengradient/account.py,sha256=2B7rtCXQDX-yF4U69h8B9-OUreJU4IqoGXG_1Hn9nWs,1150
 opengradient/cli.py,sha256=kdYR_AFKHV99HtO_son7vHpM5jWVZe8FO0iMWxJ7pJE,24444
 opengradient/client.py,sha256=RdlTz60NJKVJihYY6oVYLfNOg6RGnJbfG-2UIxUk-ws,37069
@@ -8,13 +8,13 @@ opengradient/types.py,sha256=QTEsygwT5AnIf8Dg9mexvVUe49nCo9N0pgZOOIp3trc,2214
 opengradient/utils.py,sha256=lUDPmyPqLwpZI-owyN6Rm3QvUjOn5pLN5G1QyriVm-E,6994
 opengradient/abi/inference.abi,sha256=MR5u9npZ-Yx2EqRW17_M-UnGgFF3mMEMepOwaZ-Bkgc,7040
 opengradient/llm/__init__.py,sha256=n_11WFPoU8YtGc6wg9cK6gEy9zBISf1183Loip3dAbI,62
-opengradient/llm/chat.py,sha256=Ia-K0og-DkZ1_IFgbgscn_gpLMXZX54F3fvfSrUq5M8,3849
+opengradient/llm/chat.py,sha256=F32yN1o8EvRbzZSJkUwI0-FmSVAWMuMTI9ho7wgW5hk,4470
 opengradient/proto/__init__.py,sha256=AhaSmrqV0TXGzCKaoPV8-XUvqs2fGAJBM2aOmDpkNbE,55
 opengradient/proto/infer.proto,sha256=13eaEMcppxkBF8yChptsX9HooWFwJKze7oLZNl-LEb8,1217
 opengradient/proto/infer_pb2.py,sha256=wg2vjLQCNv6HRhYuIqgj9xivi3nO4IPz6E5wh2dhDqY,3446
 opengradient/proto/infer_pb2_grpc.py,sha256=y5GYwD1EdNs892xx58jdfyA0fO5QC7k3uZOtImTHMiE,6891
-opengradient-0.3.13.dist-info/METADATA,sha256=XiQ5o2HcbeKfAX27tiHRQ9oQyJNYmkQSnmLPtl_Elz8,7612
-opengradient-0.3.13.dist-info/WHEEL,sha256=C2FUgwZgiLbznR-k0b_5k3Ai_1aASOXDss3lzCUsUug,87
-opengradient-0.3.13.dist-info/entry_points.txt,sha256=yUKTaJx8RXnybkob0J62wVBiCp_1agVbgw9uzsmaeJc,54
-opengradient-0.3.13.dist-info/licenses/LICENSE,sha256=xEcvQ3AxZOtDkrqkys2Mm6Y9diEnaSeQRKvxi-JGnNA,1069
-opengradient-0.3.13.dist-info/RECORD,,
+opengradient-0.3.15.dist-info/METADATA,sha256=j1LRy_faAg69fDQNEJhByO7kiPtd6jO-DQbNhDk99h0,8692
+opengradient-0.3.15.dist-info/WHEEL,sha256=C2FUgwZgiLbznR-k0b_5k3Ai_1aASOXDss3lzCUsUug,87
+opengradient-0.3.15.dist-info/entry_points.txt,sha256=yUKTaJx8RXnybkob0J62wVBiCp_1agVbgw9uzsmaeJc,54
+opengradient-0.3.15.dist-info/licenses/LICENSE,sha256=xEcvQ3AxZOtDkrqkys2Mm6Y9diEnaSeQRKvxi-JGnNA,1069
+opengradient-0.3.15.dist-info/RECORD,,