File tree Expand file tree Collapse file tree 1 file changed +44
-0
lines changed Expand file tree Collapse file tree 1 file changed +44
-0
lines changed Original file line number Diff line number Diff line change 1
1
# Agent for Commit0
2
2
This tool provides a command-line interface for configuring and running AI agents to assist with code development and testing.
3
3
4
+
5
+ ## (Update) Running with OpenHands
6
+
7
+ ** Step 1** : Clone (OpenHands)[ https://github.com/All-Hands-AI/OpenHands/tree/main ] and install (OpenHands)[ https://github.com/All-Hands-AI/OpenHands/blob/main/evaluation/README.md#development-environment ]
8
+
9
+ ** Step 2** : Create ` config.toml ` and write
10
+
11
+ ```
12
+ [core]
13
+ workspace_base="~/OpenHands/evaluation/benchmarks/commit0_bench"
14
+
15
+ [llm]
16
+ model="anthropic/claude-3-5-sonnet-20241022"
17
+ api_key="..."
18
+ embedding_model=""
19
+ temperature = 0.0
20
+ caching_prompt = true
21
+ ```
22
+
23
+
24
+ ** Step 3** : Run
25
+ ``` bash
26
+ ./evaluation/benchmarks/commit0_bench/scripts/run_infer.sh SPLIT MODEL HEAD CodeActAgent 16 STEPS PARALLEL_NUMBER
27
+
28
+ # Example
29
+ ./evaluation/benchmarks/commit0_bench/scripts/run_infer.sh lite llm.eval_deepseekv3 HEAD CodeActAgent 16 100 2
30
+ ```
31
+
32
+ ** Step 3.1** :
33
+ You can do the following before running code to parallelize them on remote server from OpenHands
34
+
35
+ ``` bash
36
+ export RUNTIME=remote
37
+ export SANDBOX_REMOTE_RUNTIME_API_URL=" https://runtime.eval.all-hands.dev"
38
+ export ALLHANDS_API_KEY=...
39
+ ```
40
+
41
+
42
+ ``` python
4
43
# # Quick Start
5
44
Configure an agent:
6
45
```bash
@@ -12,6 +51,11 @@ Run an agent on a specific branch:
12
51
agent run [OPTIONS] BRANCH
13
52
```
14
53
54
+ ### Example
55
+ ``` bash
56
+ agent run sonnet --max-parallel-repos 16 --agent-config-file .agent_sonnet.yaml --commit0-config-file .commit0.yaml
57
+ ```
58
+
15
59
For more detailed information on available commands and options:
16
60
``` bash
17
61
agent -h
You can’t perform that action at this time.
0 commit comments