beamds.beam.llm package#

Subpackages#

beamds.beam.llm.simulators package
- Subpackages
- Module contents

Submodules#

beamds.beam.llm.chat module#

beamds.beam.llm.conversation module#

Copied from: https://raw.githubusercontent.com/lm-sys/FastChat/main/fastchat/conversation.py

Conversation prompt templates.

We kindly request that you import fastchat instead of copying this file if you wish to use it. If you have any changes in mind, please contribute back so the community can benefit collectively and continue to maintain these valuable templates.

class beamds.beam.llm.conversation.Conversation(name: str, system_template: str = '{system_message}', system_message: str = '', roles: Tuple[str] = ('USER', 'ASSISTANT'), messages: List[List[str]] = (), offset: int = 0, sep_style: SeparatorStyle = SeparatorStyle.ADD_COLON_SINGLE, sep: str = '\n', sep2: str = None, stop_str: str | List[str] = None, stop_token_ids: List[int] = None, max_image_size_mb: int = None)[source]#

Bases: object

A class that manages prompt templates and keeps all conversation history.

append_message(role: str, message: str)[source]#: Append a new message.

convert_image_to_base64(image)[source]#: Given an image, return the base64 encoded image string.

copy()[source]#

dict()[source]#

extract_text_and_image_hashes_from_messages()[source]#

get_images()[source]#

get_prompt() → str[source]#: Get the prompt for generation.

get_system_message()[source]#: return the system message.

max_image_size_mb: int = None#

messages: List[List[str]] = ()#

name: str#

offset: int = 0#

roles: Tuple[str] = ('USER', 'ASSISTANT')#

save_new_images(has_csam_images=False, use_remote_storage=False)[source]#

sep: str = '\n'#

sep2: str = None#

sep_style: SeparatorStyle = 1#

set_system_message(system_message: str)[source]#: Set the system message.

stop_str: str | List[str] = None#

stop_token_ids: List[int] = None#

system_message: str = ''#

system_template: str = '{system_message}'#

to_anthropic_vision_api_messages()[source]#: Convert the conversation to Claude-3 Messages Vision API format

to_gemini_api_messages()[source]#

to_gradio_chatbot()[source]#: Convert the conversation to gradio chatbot format.

to_openai_api_messages()[source]#: Convert the conversation to OpenAI chat completion format.

to_openai_image_format(image_urls)[source]#

to_openai_vision_api_messages()[source]#: Convert the conversation to OpenAI vision api completion format

to_reka_api_messages()[source]#

to_vertex_api_messages()[source]#

update_last_message(message: str)[source]#

Update the last output.

The last message is typically set to be None when constructing the prompt, so we need to update it in-place after getting the response from a model.

class beamds.beam.llm.conversation.SeparatorStyle(value, names=<not given>, *values, module=None, qualname=None, type=None, start=1, boundary=None)[source]#

Bases: IntEnum

Separator styles.

ADD_COLON_SINGLE = 1#

ADD_COLON_SPACE_SINGLE = 3#

ADD_COLON_TWO = 2#

ADD_NEW_LINE_SINGLE = 6#

CHATGLM = 9#

CHATGLM3 = 17#

CHATINTERN = 11#

CHATML = 10#

CLLM = 22#

DEEPSEEK_CHAT = 18#

DEFAULT = 23#

DOLLY = 12#

FALCON_CHAT = 16#

GEMMA = 21#

LLAMA2 = 7#

LLAMA3 = 8#

METAMATH = 19#

NO_COLON_SINGLE = 4#

NO_COLON_TWO = 5#

PHOENIX = 14#

ROBIN = 15#

RWKV = 13#

YUAN2 = 20#

beamds.beam.llm.conversation.get_conv_template(name: str) → Conversation[source]#: Get a conversation template.

beamds.beam.llm.conversation.register_conv_template(template: Conversation, override: bool = False)[source]#: Register a new conversation template.

beamds.beam.llm.core module#

beamds.beam.llm.hf_conversation module#

class beamds.beam.llm.hf_conversation.Conversation(messages: str | List[Dict[str, str]] = None, conversation_id: UUID = None, **deprecated_kwargs)[source]#

Bases: object

Utility class containing a conversation and its history. This class is meant to be used as an input to the [ConversationalPipeline]. The conversation contains several utility functions to manage the addition of new user inputs and generated model responses.

Parameters:

messages (Union[str, List[Dict[str, str]]], optional) – The initial messages to start the conversation, either a string, or a list of dicts containing “role” and “content” keys. If a string is passed, it is interpreted as a single message with the “user” role.
conversation_id (uuid.UUID, optional) – Unique identifier for the conversation. If not provided, a random UUID4 id will be assigned to the conversation.

Usage:

`python conversation = Conversation("Going to the movies tonight - any suggestions?") conversation.add_message({"role": "assistant", "content": "The Big lebowski."}) conversation.add_message({"role": "user", "content": "Is it good?"}) `

add_message(message: Dict[str, str])[source]#

add_user_input(text: str, overwrite: bool = False)[source]#: Add a user input to the conversation for the next round. This is a legacy method that assumes that inputs must alternate user/assistant/user/assistant, and so will not add multiple user messages in succession. We recommend just using add_message with role “user” instead.

append_response(response: str)[source]#: This is a legacy method. We recommend just using add_message with an appropriate role instead.

property generated_responses#

iter_texts()[source]#

mark_processed()[source]#: This is a legacy method, as the Conversation no longer distinguishes between processed and unprocessed user input. We set a counter here to keep behaviour mostly backward-compatible, but in general you should just read the messages directly when writing new code.

property new_user_input#

property past_user_inputs#

beamds.beam.llm.model_adapter module#

Copied from: https://raw.githubusercontent.com/lm-sys/FastChat/main/fastchat/model/model_adapter.py removed the transformer dependency and add local conversation dependency remove functions: get_generate_stream_function, load_model remove imports: psutils, math, torch

Model adapter registration.

class beamds.beam.llm.model_adapter.AiroborosAdapter[source]#

beamds.beam.llm package#

Subpackages#

Submodules#

beamds.beam.llm.chat module#

beamds.beam.llm.conversation module#

beamds.beam.llm.core module#

beamds.beam.llm.hf_conversation module#

beamds.beam.llm.model_adapter module#

beamds.beam.llm.models module#

beamds.beam.llm.openai module#

beamds.beam.llm.resource module#

beamds.beam.llm.response module#

beamds.beam.llm.task module#

beamds.beam.llm.tools module#

beamds.beam.llm.utils module#

Module contents#

This Page