update

Author: Yunnglin

cookbook/client/tinker/self_host/sample.py

@@ -52,7 +52,6 @@
 params = types.SamplingParams(
     max_tokens=128,       # Maximum number of tokens to generate
     temperature=0.7,
-    stop=['\n']          # Stop generation when a newline character is produced
 )
 
 # Step 6: Send the sampling request to the server.

cookbook/client/twinkle/self_host/self_congnition.py

@@ -31,6 +31,11 @@
 # - api_key: authentication token (loaded from environment variable)
 client = init_twinkle_client(base_url=base_url, api_key=api_key)
 
+# List available models of the server
+print('Available models:')
+for item in client.get_server_capabilities().supported_models:
+    print('- ' + item.model_name)
+
 # Step 3: Query the server for existing training runs and their checkpoints.
 # This is useful for resuming a previous training session.
 runs = client.list_training_runs()

src/twinkle_client/manager.py

@@ -5,7 +5,7 @@
 import threading
 from typing import Any, Dict, List, Optional, Tuple
 from twinkle import get_logger
-from twinkle_client.types.server import (DeleteCheckpointResponse, GetServerCapabilitiesResponse, SupportedModel)
+from twinkle_client.types.server import (DeleteCheckpointResponse, GetServerCapabilitiesResponse)
 from twinkle_client.types.session import (CreateSessionRequest, CreateSessionResponse, SessionHeartbeatRequest,
                                            SessionHeartbeatResponse)
 from twinkle_client.types.training import (Checkpoint, Cursor, ParsedCheckpointTwinklePath, TrainingRun,