Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Code Generation on ClassEval
Loading...
27
Pass@1 (Class-level)
DeepSeek-Coder-6.7B-Instruct
8.28
13.14
18
22.86
Oct 31, 2024
Pass@1 (Class-level)
Pass@1 (Method-level)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Pass@1 (Class-level)
Pass@1 (Method-level)
DeepSeek-Coder-6.7B-Instruct
decoding_strategy=greedy
2024.10
27
57.2
CodeQwen1.5-7B-Chat
decoding_strategy=greedy
2024.10
27
54.6
SelfCodeAlign-CQ-7B
decoding_strategy=greedy
2024.10
27
52.6
Llama-3-8B-Instruct
decoding_strategy=greedy
2024.10
23
52.4
CodeQwen1.5-7B-Base
decoding_strategy=greedy
2024.10
23
52.8
CodeGemma-7B-IT
decoding_strategy=greedy
2024.10
21
44.8
CodeQwen1.5-7B-OctoPack
decoding_strategy=greedy
2024.10
21
45.2
OctoCoder-16B
decoding_strategy=greedy
2024.10
19
38
StarCoder2-15B
decoding_strategy=greedy
2024.10
9
24.9
Feedback
Search any
task
Search any
task