Yehowshua/ZIm
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Agent Eval: Distinguish tool successes and failures in log (#28839)

Browse source 
Release Notes:

- N/A
This commit is contained in:
Michael Sloan 2025-04-15 22:51:33 -06:00 • committed by GitHub
parent 6267a147ba
commit da2d8bd845
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 19 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

30
crates/eval/src/example.rs
View file

@ -375,18 +375,26 @@ impl Example {
pending_tool_use,
..
} => {
if let Some(tool_use) = pending_tool_use {
let message = format!("TOOL FINISHED: {}", tool_use.name);
println!("{}{message}", log_prefix);
writeln!(&mut output_file, "\n{}", message).log_err();
}
thread.update(cx, |thread, _cx| {
if let Some(tool_result) = thread.tool_result(&tool_use_id) {
writeln!(&mut output_file, "\n{}\n", tool_result.content).log_err();
let mut tool_use_counts = tool_use_counts.lock().unwrap();
*tool_use_counts
.entry(tool_result.tool_name.clone())
.or_insert(0) += 1;
if let Some(tool_use) = pending_tool_use {
if let Some(tool_result) = thread.tool_result(&tool_use_id) {
let message = if tool_result.is_error {
format!("TOOL FAILED: {}", tool_use.name)
} else {
format!("TOOL FINISHED: {}", tool_use.name)
};
println!("{log_prefix}{message}");
writeln!(&mut output_file, "\n{}", message).log_err();
writeln!(&mut output_file, "\n{}\n", tool_result.content).log_err();
let mut tool_use_counts = tool_use_counts.lock().unwrap();
*tool_use_counts
.entry(tool_result.tool_name.clone())
.or_insert(0) += 1;
} else {
let message = format!("TOOL FINISHED WITHOUT RESULT: {}", tool_use.name);
println!("{log_prefix}{message}");
writeln!(&mut output_file, "\n{}", message).log_err();
}
}
})?;
}