summaryrefslogtreecommitdiff
path: root/help.py
diff options
context:
space:
mode:
authorSchark <jordan@schark.online>2023-12-01 23:11:46 -0800
committerSchark <jordan@schark.online>2023-12-01 23:11:46 -0800
commitf580735aaafdc8e0c9128955fcb93c039d10b8b9 (patch)
tree604b2179c88bb419ec0421ae2c4a76c67518ae28 /help.py
downloadcli-gpt-f580735aaafdc8e0c9128955fcb93c039d10b8b9.tar.gz
cli-gpt-f580735aaafdc8e0c9128955fcb93c039d10b8b9.zip
Initial work
Diffstat (limited to '')
-rw-r--r--help.py155
1 files changed, 155 insertions, 0 deletions
diff --git a/help.py b/help.py
new file mode 100644
index 0000000..913ec6f
--- /dev/null
+++ b/help.py
@@ -0,0 +1,155 @@
+import json
+import subprocess
+from pathlib import Path
+
+import tiktoken
+
+# TODO: I don't love the way this file is structured from an extensibility perspective.
+# Find a way to maybe like the 'options' dict with the command list?
+
+def start_chat(model: str):
+ print("System: Welcome to cli-gpt. You may type your questions, or seek additional functionality via '/help'.")
+ print(f"System: Currently using model '{model}'.")
+ return [{"role": "system", "content": "You are a helpful assistant."}]
+
+def get_token_count(messages: list, model: str):
+ total_tokens: int = 0
+ encoding = tiktoken.encoding_for_model(model)
+ for message in messages:
+ if message.get("role") == "system":
+ continue
+ text = message.get("content", "")
+ message_tokens = len(encoding.encode(text))
+ total_tokens += message_tokens
+ return total_tokens
+
+class HelpCommands:
+
+ options: dict = {
+ "/exit": "Closes the chat.",
+ "/context": "Passthrough a URL to curl the context of into the chat history.",
+ "/help": "Display this list of available commands.",
+ "/load": "Load in a previous chat's JSON file.",
+ "/save": "Saves messages to specified JSON file and closes chat.",
+ "/clear": "Clears all messages and tokens from the chatlog, restarting the chat.",
+ "/model": "Change the model being used.",
+ "/info": "Print model information and cli-gpt version.",
+ }
+
+ text_models = [
+ "gpt-3.5-turbo",
+ "gpt-4",
+ "gpt-4-32k",
+ "gpt-4-0613",
+ "gpt-4-32k-0613",
+ "gpt-4-32k-0613",
+ "gpt-4-1106-preview",
+ "gpt-4-vision-preview",
+ ]
+
+ image_models = [
+ "dall-e-2",
+ "dall-e-3",
+ ]
+
+ model_type = None
+
+ def __init__(self, model: str):
+ if model in self.text_models:
+ self.model_type = "text"
+ elif model in self.image_models:
+ self.model_type = "image"
+ else:
+ raise TypeError(f"System: Model '{model}' is not available. Please start again and re-specify model version, or leave blank.")
+
+
+ def command(self, user_input: str, messages: list, model: str) -> (int, list, str):
+ if user_input.lower() in list(self.options.keys()):
+ user_input_lower = user_input.lower()
+
+ if user_input_lower == "/exit":
+ return 1, [None], ""
+
+ if user_input_lower == "/context":
+ print("\nSystem: Please provide the URL you would like to curl.")
+ url = input("URL: ")
+ curl_output = subprocess.check_output(f"curl {url}", shell=True).decode('utf-8').strip()
+ return [None, f"I would like to provide the following context from a curl command to '{url} to this chat: {curl_output}"]
+
+ if user_input_lower == "/help":
+ print(f"\nSystem: Below is a list of available commands.\n")
+ for key in list(self.options.keys()):
+ print(f" - {key}: {self.options[key]}")
+ return 2, messages, model
+
+ if user_input_lower == "/load":
+ print("\nSystem: Please specify the filepath you would like to load in from, or '/cancel'.")
+ url = input("Path: ")
+ if url != "/cancel":
+ with open(Path(url), "r") as file:
+ messages = json.load(file)
+ print(f"System: Successfully read in from {url}. Continuing chat.")
+ return 2, messages, model
+
+ if user_input_lower == "/save":
+ status, destination = self._save(messages)
+ if status == 1:
+ return 2, messages, model
+ print(f"System: Successfully saved to {destination}. Closing chat.")
+ return 1, [None], ""
+
+ if user_input_lower == "/clear":
+ print(f"\nSystem: Clearing messages and restarting log.\n\n")
+ messages = start_chat(model)
+ return 2, messages, model
+
+ if user_input_lower == "/info":
+ print(f"\nSystem: This chatlog has used {get_token_count(messages, model)} token for model version '{model}'.")
+ print("System: Currently using cli-gpt version 0.0.1.")
+ return 2, messages, model
+
+ if user_input_lower == "/model":
+ print("\nSystem: Below is a list of available models. View up-to-date model information in the OpenAI API documentation.")
+ print("\n - Text Models")
+ for list_model in self.text_models:
+ print(f" - {list_model}")
+ print("\n - Image Models")
+ for list_model in self.image_models:
+ print(f" - {list_model}")
+ print("\nSystem: Change model version below, use '/list' to reprint available models, or '/cancel' to return to chat.")
+ new_model = input("\nModel: ")
+ while new_model not in self.text_models and new_model not in self.image_models:
+ if new_model == "/list":
+ print("\nSystem: Below is a list of available models. View up-to-date model information in the OpenAI API documentation.")
+ print("\n - Text Models")
+ for list_model in self.text_models:
+ print(f" - {list_model}")
+ print("\n - Image Models")
+ for list_model in self.image_models:
+ print(f" - {list_model}")
+ elif new_model == "/cancel":
+ return 2, messages, model
+ else:
+ print(f"System: '{new_model}' is not an accepted model. Use 'list' to output available models.")
+ new_model = input("\nModel: ")
+
+ # "image" models and "text" models behave different, handle switching
+ if (self.model_type == "text" and new_model in self.image_models) or (self.model_type == "image" and new_model in self.text_models):
+ print("\nSystem: Switching between 'text' and 'image' models requires clearing the current message log. Would you like to save before switching models?")
+ user_save = input("Save? (y,N): ")
+ if user_save.lower() == "y":
+ self._save(messages)
+ return 2, messages, new_model
+
+ messages.append({"role": "user", "content": user_input})
+ return 0, messages, model
+
+ def _save(self, messages):
+ print("\nSystem: Please specify the filepath you would like to save to as a *.json file, or '/cancel'.")
+ url = input("Path: ")
+ if url == "/cancel":
+ return 1, None
+ with open(Path(url), "w+") as file:
+ json.dump(messages, file)
+ return 0, url
+