dearwolf/aider
1
0
Fork 0
mirror of https://github.com/Aider-AI/aider.git synced 2025-05-31 01:35:00 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 9

copy

Browse source
This commit is contained in:
Paul Gauthier 2025-05-08 11:41:51 -07:00
parent d5ea078f24
commit acd7309b78
2 changed files with 13 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

14
aider/website/_data/qwen3_leaderboard.yml
View file

@ -1,6 +1,6 @@
- dirname: 2025-05-08-03-20-24--qwen3-32b-default
test_cases: 225
model: Qwen3 32B on OpenRouter, all providers, default settings (thinking)
model: Qwen3 32B diff on OpenRouter, all providers, default settings (thinking)
edit_format: diff
commit_hash: aaacee5-dirty, aeaf259
pass_rate_1: 14.2
@ -28,7 +28,7 @@
- dirname: 2025-05-08-03-22-37--qwen3-235b-defaults
test_cases: 225
model: Qwen3 235B A22B on OpenRouter, all providers, default settings (thinking)
model: Qwen3 235B A22B diff on OpenRouter, all providers, default settings (thinking)
edit_format: diff
commit_hash: aaacee5-dirty
pass_rate_1: 17.3
@ -57,7 +57,7 @@
- dirname: 2025-05-08-17-39-14--qwen3-235b-or-together-only
test_cases: 225
model: Qwen3 235B A22B on OpenRouter only TogetherAI, recommended /no_think settings
model: Qwen3 235B A22B diff on OpenRouter only TogetherAI, recommended /no_think settings
edit_format: diff
commit_hash: 328584e
pass_rate_1: 28.0
@ -86,7 +86,7 @@
- dirname: 2025-04-30-04-49-37--Qwen3-235B-A22B-whole-nothink
test_cases: 225
model: Qwen3-235B-A22B with VLLM, bfloat16, recommended /no_think settings
model: Qwen3-235B-A22B whole with VLLM, bfloat16, recommended /no_think settings
edit_format: whole
commit_hash: 0c383df-dirty
pass_rate_1: 28.0
@ -112,7 +112,7 @@
- dirname: 2025-04-30-04-49-50--Qwen3-235B-A22B-diff-nothink
test_cases: 225
model: Qwen3-235B-A22B with VLLM, bfloat16, recommended /no_think settings
model: Qwen3-235B-A22B diff with VLLM, bfloat16, recommended /no_think settings
edit_format: diff
commit_hash: 0c383df-dirty
pass_rate_1: 29.8
@ -138,7 +138,7 @@
- dirname: 2025-04-30-04-08-41--Qwen3-32B-whole-nothink
test_cases: 225
model: Qwen3-32B with VLLM, bfloat16, recommended /no_think settings
model: Qwen3-32B whole with VLLM, bfloat16, recommended /no_think settings
edit_format: whole
commit_hash: 0c383df-dirty
pass_rate_1: 20.4
@ -164,7 +164,7 @@
- dirname: 2025-04-30-04-08-51--Qwen3-32B-diff-nothink
test_cases: 225
model: Qwen3-32B with VLLM, bfloat16, recommended /no_think settings
model: Qwen3-32B diff with VLLM, bfloat16, recommended /no_think settings
edit_format: diff
commit_hash: 0c383df-dirty
pass_rate_1: 20.4

9
aider/website/_posts/2025-05-08-qwen3.md
View file

@ -14,10 +14,13 @@ Open source models are wonderful because anyone can serve them,
but API providers can use very different inference settings, quantizations, etc.
Below are collection of aider polyglot benchmark results for the new Qwen3 models.
Results are presented with various settings against various API providers,
with the hope of showcasing the strengths of these models and their providers.
Results are presented using both "diff" and "whole"
[edit formats](https://aider.chat/docs/more/edit-formats.html),
with various models settings, against various API providers.
See details for configuring Qwen3 after the results table.
See details on the
[model settings](https://aider.chat/docs/config/adv-model-settings.html#model-settings)
used after the results table.
{: .note }
This article is being updated as new results become available.