Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

A few fixes for TAC evaluation harness #6586

Merged
merged 10 commits into from
Feb 15, 2025
Merged

A few fixes for TAC evaluation harness #6586

Changes from 3 commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
7bb4461
Enable enable_som_visual_browsing in TAC evaluation harness
li-boxuan Feb 2, 2025
078c275
Log set of marks
li-boxuan Feb 2, 2025
37b51f4
Disable github microagent
li-boxuan Feb 2, 2025
348f69d
Fix path in run_infer.sh
li-boxuan Feb 2, 2025
a644190
Fix python path
li-boxuan Feb 2, 2025
45ea2b1
Disable prompt extensions
li-boxuan Feb 2, 2025
5cc30cd
Fix for base64 header - might be caused by upgrade of browsergym
li-boxuan Feb 2, 2025
aae611a
Check existence before pull
li-boxuan Feb 3, 2025
22a9f55
Merge remote-tracking branch 'upstream/main' into boxuanli/eval-visua…
li-boxuan Feb 14, 2025
3c90538
Fix signature
li-boxuan Feb 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 18 additions & 1 deletion evaluation/benchmarks/the_agent_company/run_infer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@
get_llm_config_arg,
get_parser,
)
from openhands.core.config.agent_config import AgentConfig
from openhands.core.logger import openhands_logger as logger
from openhands.core.main import create_runtime, run_controller
from openhands.events.action import CmdRunAction, MessageAction
Expand All @@ -34,6 +35,7 @@ def get_config(
task_short_name: str,
mount_path_on_host: str,
llm_config: LLMConfig,
agent_config: AgentConfig,
) -> AppConfig:
config = AppConfig(
run_as_openhands=False,
Expand All @@ -57,6 +59,7 @@ def get_config(
workspace_mount_path_in_sandbox='/outputs',
)
config.set_llm_config(llm_config)
config.set_agent_config(agent_config)
return config


Expand Down Expand Up @@ -152,6 +155,12 @@ def run_solver(
os.path.join(screenshots_dir, f'{image_id}.png'), 'wb'
) as file:
file.write(image_data)
if obs.set_of_marks:
som_image_data = base64.b64decode(obs.set_of_marks)
with open(
os.path.join(screenshots_dir, f'{image_id}_som.png'), 'wb'
) as file:
file.write(som_image_data)

if save_final_state:
os.makedirs(state_dir, exist_ok=True)
Expand Down Expand Up @@ -214,6 +223,14 @@ def run_evaluator(
)
args, _ = parser.parse_known_args()

#### TODO: parse this from cli and toml ####
agent_config: AgentConfig | None = None
agent_config = AgentConfig(
enable_som_visual_browsing=True,
disabled_microagents=['github'],
enyst marked this conversation as resolved.
Show resolved Hide resolved
)
############################################

agent_llm_config: LLMConfig | None = None
if args.agent_llm_config:
agent_llm_config = get_llm_config_arg(args.agent_llm_config)
Expand Down Expand Up @@ -254,7 +271,7 @@ def run_evaluator(
else:
temp_dir = tempfile.mkdtemp()
config: AppConfig = get_config(
args.task_image_name, task_short_name, temp_dir, agent_llm_config
args.task_image_name, task_short_name, temp_dir, agent_llm_config, agent_config
)
runtime: Runtime = create_runtime(config)
call_async_from_sync(runtime.connect)
Expand Down