diff --git a/.github/workflows/code_format.yml b/.github/workflows/code_format.yml index d0623472..5542a395 100644 --- a/.github/workflows/code_format.yml +++ b/.github/workflows/code_format.yml @@ -21,7 +21,7 @@ jobs: - name: Set up Python uses: actions/setup-python@v5 with: - python-version: '3.10' + python-version: '3.11' cache: 'pip' # caching pip dependencies - name: Pip install @@ -31,4 +31,4 @@ jobs: run: pip list - name: Code Formatting - run: black . --check + run: black . --check --diff diff --git a/src/agentlab/agents/generic_agent/__init__.py b/src/agentlab/agents/generic_agent/__init__.py index af5d9b57..f6cbc89c 100644 --- a/src/agentlab/agents/generic_agent/__init__.py +++ b/src/agentlab/agents/generic_agent/__init__.py @@ -15,7 +15,10 @@ RANDOM_SEARCH_AGENT, AGENT_4o, AGENT_4o_MINI, + AGENT_CLAUDE_SONNET_35, AGENT_4o_VISION, + AGENT_4o_MINI_VISION, + AGENT_CLAUDE_SONNET_35_VISION, ) __all__ = [ @@ -28,4 +31,8 @@ "AGENT_8B", "RANDOM_SEARCH_AGENT", "AGENT_CUSTOM", + "AGENT_CLAUDE_SONNET_35", + "AGENT_4o_VISION", + "AGENT_4o_MINI_VISION", + "AGENT_CLAUDE_SONNET_35_VISION", ] diff --git a/src/agentlab/agents/generic_agent/agent_configs.py b/src/agentlab/agents/generic_agent/agent_configs.py index 86f617da..02a4c7dc 100644 --- a/src/agentlab/agents/generic_agent/agent_configs.py +++ b/src/agentlab/agents/generic_agent/agent_configs.py @@ -260,6 +260,10 @@ chat_model_args=CHAT_MODEL_ARGS_DICT["openai/gpt-4o-mini-2024-07-18"], flags=FLAGS_GPT_4o, ) +AGENT_CLAUDE_SONNET_35 = GenericAgentArgs( + chat_model_args=CHAT_MODEL_ARGS_DICT["openrouter/anthropic/claude-3.5-sonnet:beta"], + flags=FLAGS_GPT_4o, +) # GPT-4o vision default config FLAGS_GPT_4o_VISION = FLAGS_GPT_4o.copy() @@ -271,6 +275,16 @@ flags=FLAGS_GPT_4o_VISION, ) +AGENT_4o_MINI_VISION = GenericAgentArgs( + chat_model_args=CHAT_MODEL_ARGS_DICT["openai/gpt-4o-mini-2024-07-18"], + flags=FLAGS_GPT_4o_VISION, +) + +AGENT_CLAUDE_SONNET_35_VISION = GenericAgentArgs( + chat_model_args=CHAT_MODEL_ARGS_DICT["openrouter/anthropic/claude-3.5-sonnet:beta"], + flags=FLAGS_GPT_4o_VISION, +) + DEFAULT_RS_FLAGS = GenericPromptFlags( flag_group="default_rs", diff --git a/src/agentlab/agents/generic_agent/reproducibility_agent.py b/src/agentlab/agents/generic_agent/reproducibility_agent.py index 19cbc5c6..d9fa4c29 100644 --- a/src/agentlab/agents/generic_agent/reproducibility_agent.py +++ b/src/agentlab/agents/generic_agent/reproducibility_agent.py @@ -5,7 +5,7 @@ This module contains the classes and functions to reproduce the results of a study. It is used to create a new study that will run the same experiments as the original study, but with a reproducibility agent that will mimic the same -answers as the original agent. +answers as the original agent. Stats are collected to compare the original agent's answers with the new agent's answers. Load the this reproducibility study in agent-xray to compare the results. diff --git a/src/agentlab/llm/llm_configs.py b/src/agentlab/llm/llm_configs.py index 6a60e157..725d8a0c 100644 --- a/src/agentlab/llm/llm_configs.py +++ b/src/agentlab/llm/llm_configs.py @@ -1,5 +1,9 @@ -from agentlab.llm.chat_api import (AzureModelArgs, OpenAIModelArgs, - OpenRouterModelArgs, SelfHostedModelArgs) +from agentlab.llm.chat_api import ( + AzureModelArgs, + OpenAIModelArgs, + OpenRouterModelArgs, + SelfHostedModelArgs, +) default_oss_llms_args = { "n_retry_server": 4,