Files
oak-gpui/crates/language/Cargo.toml
T
Oleksiy Syvokon ab017129d8 agent: Improve Gemini support in the edit_file tool (#31116)
This change improves `eval_extract_handle_command_output` results for
all models:

Model                       | Pass rate before | Pass rate after
----------------------------|------------------|----------------
claude-3.7-sonnet           |  0.96            | 0.98
gemini-2.5-pro              |  0.35            | 0.86
gpt-4.1                     |  0.81            | 1.00

Part of this improvement comes from more robust evaluation, which now
accepts multiple possible outcomes. Another part is from the prompt
adaptation: addressing common Gemini failure modes, adding a few-shot
example, and, in the final commit, auto-rewriting instructions for
clarity and conciseness.

This change still needs validation from larger end-to-end evals.


Release Notes:

- N/A
2025-05-22 12:01:43 +03:00

95 lines
2.5 KiB
TOML

[package]
name = "language"
version = "0.1.0"
edition.workspace = true
publish.workspace = true
license = "GPL-3.0-or-later"
[lints]
workspace = true
[lib]
path = "src/language.rs"
doctest = false
[features]
test-support = [
"rand",
"collections/test-support",
"lsp/test-support",
"text/test-support",
"tree-sitter-rust",
"tree-sitter-python",
"tree-sitter-typescript",
"settings/test-support",
"util/test-support",
]
[dependencies]
anyhow.workspace = true
async-trait.workspace = true
async-watch.workspace = true
clock.workspace = true
collections.workspace = true
ec4rs.workspace = true
fs.workspace = true
futures.workspace = true
fuzzy.workspace = true
globset.workspace = true
gpui.workspace = true
http_client.workspace = true
imara-diff.workspace = true
itertools.workspace = true
log.workspace = true
lsp.workspace = true
parking_lot.workspace = true
postage.workspace = true
rand = { workspace = true, optional = true }
regex.workspace = true
rpc.workspace = true
schemars.workspace = true
serde.workspace = true
serde_json.workspace = true
settings.workspace = true
smallvec.workspace = true
smol.workspace = true
streaming-iterator.workspace = true
strsim.workspace = true
sum_tree.workspace = true
task.workspace = true
text.workspace = true
theme.workspace = true
tree-sitter-python = { workspace = true, optional = true }
tree-sitter-rust = { workspace = true, optional = true }
tree-sitter-typescript = { workspace = true, optional = true }
tree-sitter.workspace = true
unicase = "2.6"
util.workspace = true
workspace-hack.workspace = true
diffy = "0.4.2"
[dev-dependencies]
collections = { workspace = true, features = ["test-support"] }
ctor.workspace = true
env_logger.workspace = true
gpui = { workspace = true, features = ["test-support"] }
indoc.workspace = true
lsp = { workspace = true, features = ["test-support"] }
pretty_assertions.workspace = true
rand.workspace = true
settings = { workspace = true, features = ["test-support"] }
text = { workspace = true, features = ["test-support"] }
http_client = { workspace = true, features = ["test-support"] }
tree-sitter-elixir.workspace = true
tree-sitter-embedded-template.workspace = true
tree-sitter-heex.workspace = true
tree-sitter-html.workspace = true
tree-sitter-json.workspace = true
tree-sitter-md.workspace = true
tree-sitter-python.workspace = true
tree-sitter-ruby.workspace = true
tree-sitter-rust.workspace = true
tree-sitter-typescript.workspace = true
unindent.workspace = true
util = { workspace = true, features = ["test-support"] }