launch_scientist.py

code:why and how are these LLM picked up?.py

try:

print_time()

print(f"*Starting idea: {idea_name}*")

io = InputOutput(yes=True, chat_history_file=f"{folder_name}/{idea_name}_aider.txt")

⭐if model == "hybrid":

main_model = Model("claude-3-5-sonnet-20240620")

elif model == "deepseek-coder-v2-0724":

main_model = Model("deepseek/deepseek-coder")

elif model == "llama3.1-405b":

main_model = Model("openrouter/meta-llama/llama-3.1-405b-instruct")

else:

main_model = Model(model)

coder = Coder.create(

main_model=main_model, fnames=fnames, io=io, stream=False, use_git=False, edit_format="diff"

)

coder is Aider's API?

code:coder.py

coder = Coder.create(

main_model=main_model, fnames=fnames, io=io, stream=False, use_git=False,

edit_format="diff"

)

generate_idea

code:.py

if __name__ == "__main__":

args = parse_arguments()

# Check available GPUs and adjust parallel processes if necessary

available_gpus = get_available_gpus(args.gpus)

if args.parallel > len(available_gpus):

print(

f"Warning: Requested {args.parallel} parallel processes, but only {len(available_gpus)} GPUs available. Adjusting to {len(available_gpus)}."

)

args.parallel = len(available_gpus)

print(f"Using GPUs: {available_gpus}")

# Create client

if args.model == "claude-3-5-sonnet-20240620":

import anthropic

print(f"Using Anthropic API with model {args.model}.")

client_model = "claude-3-5-sonnet-20240620"

client = anthropic.Anthropic()

elif args.model == "gpt-4o-2024-05-13" or args.model == "hybrid":

import openai

print(f"Using OpenAI API with model {args.model}.")

client_model = "gpt-4o-2024-05-13"

client = openai.OpenAI()

elif args.model == "deepseek-coder-v2-0724":

import openai

print(f"Using OpenAI API with {args.model}.")

client_model = "deepseek-coder-v2-0724"

client = openai.OpenAI(

)

elif args.model == "llama3.1-405b":

import openai

print(f"Using OpenAI API with {args.model}.")

client_model = "meta-llama/llama-3.1-405b-instruct"

client = openai.OpenAI(

)

else:

raise ValueError(f"Model {args.model} not supported.")

base_dir = osp.join("templates", args.experiment)

results_dir = osp.join("results", args.experiment)

ideas = ⭐️generate_ideas(

base_dir,

client=client,

model=client_model,

skip_generation=args.skip_idea_generation,

max_num_generations=args.num_ideas,

num_reflections=NUM_REFLECTIONS,

)

ideas = ⭐️check_idea_novelty(

ideas,

base_dir=base_dir,

client=client,

model=client_model,

)

with open(osp.join(base_dir, "ideas.json"), "w") as f:

json.dump(ideas, f, indent=4)

# novel_ideas = list(reversed(novel_ideas))

if args.parallel > 0:

print(f"Running {args.parallel} parallel processes")

queue = multiprocessing.Queue()

for idea in novel_ideas:

queue.put(idea)

processes = []

for i in range(args.parallel):

p = multiprocessing.Process(

target=worker,

args=(

queue,

base_dir,

results_dir,

args.model,

client,

client_model,

args.writeup,

args.improvement,

gpu_id,

)

)

p.start()

time.sleep(150)

processes.append(p)

# Signal workers to exit

for _ in range(args.parallel):

queue.put(None)

for p in processes:

p.join()

print("All parallel processes completed.")

else:

for idea in novel_ideas:

try:

success = ⭐️do_idea(

base_dir,

results_dir,

idea,

args.model,

client,

client_model,

args.writeup,

args.improvement,

)

except Exception as e:

print("All ideas evaluated.")