run-llama/llama-lab
1
0
Fork 0
mirror of https://github.com/run-llama/llama-lab.git synced 2026-07-01 20:14:02 -04:00
Code Issues 12 Packages Projects Releases Wiki Activity

add max iterations to streamlit runner

Browse Source
This commit is contained in:
Logan Markewich
2023-04-23 22:17:31 -06:00
parent e117625c8b
commit 0262d9d005
1 changed files with 49 additions and 36 deletions
Download Patch File Download Diff File Expand all files Collapse all files
llama_agi/llama_agi/runners/AutoStreamlitAGIRunner.py
+49 -36
View File
@@ -27,6 +27,7 @@ class AutoStreamlitAGIRunner(BaseAGIRunner):
initial_task: str,
sleep_time: int,
initial_task_list: Optional[List[str]] = None,
max_iterations: Optional[int] = None
) -> None:
logs_col, state_col = st.columns(2)
@@ -77,47 +78,59 @@ class AutoStreamlitAGIRunner(BaseAGIRunner):
if state_str:
st_state.markdown(state_str.replace("\n", "\n\n"))
iteration = 0
paused = False
while True:
# Get the next task
cur_task = self.task_manager.get_next_task()
if paused and st.button(f"Continue for {max_iterations} more steps?"):
paused = False
st.success("Continuing..")
else:
iteration += 1
# Execute current task
result_dict = self.execution_agent.execute_task(
objective=objective,
cur_task=cur_task,
completed_tasks_summary=completed_tasks_summary,
)
result = result_dict['output']
# update logs
log = make_intermediate_steps_pretty(json.dumps(result_dict['intermediate_steps'])) + [result]
logs.append(log)
st_logs.write(log)
# Get the next task
cur_task = self.task_manager.get_next_task()
# store the task and result as completed
self.task_manager.add_completed_task(cur_task, result)
# Execute current task
result_dict = self.execution_agent.execute_task(
objective=objective,
cur_task=cur_task,
completed_tasks_summary=completed_tasks_summary,
)
result = result_dict['output']
# update logs
log = make_intermediate_steps_pretty(json.dumps(result_dict['intermediate_steps'])) + [result]
logs.append(log)
st_logs.write(log)
# generate new task(s), if needed
self.task_manager.generate_new_tasks(objective, cur_task, result)
# store the task and result as completed
self.task_manager.add_completed_task(cur_task, result)
# Summarize completed tasks
completed_tasks_summary = self.task_manager.get_completed_tasks_summary()
# generate new task(s), if needed
self.task_manager.generate_new_tasks(objective, cur_task, result)
# log state of AGI to streamlit
state_str = log_current_status(
cur_task,
result,
completed_tasks_summary,
self.task_manager.current_tasks,
return_str=True
)
if state_str is not None:
st_state.markdown(state_str.replace("\n", "\n\n"))
# Summarize completed tasks
completed_tasks_summary = self.task_manager.get_completed_tasks_summary()
# Quit the loop?
if len(self.task_manager.current_tasks) == 0:
print("Out of tasks! Objective Accomplished?")
break
# log state of AGI to streamlit
state_str = log_current_status(
cur_task,
result,
completed_tasks_summary,
self.task_manager.current_tasks,
return_str=True
)
if state_str is not None:
st_state.markdown(state_str.replace("\n", "\n\n"))
# wait a bit to let you read what's happening
time.sleep(sleep_time)
# Quit the loop?
if len(self.task_manager.current_tasks) == 0:
st.success("Out of tasks! Objective Accomplished?")
break
if max_iterations is not None and iteration % max_iterations == 0:
st.warning("Reached max iterations, pausing!")
paused = True
# wait a bit to let you read what's happening
time.sleep(sleep_time)