Update README.md
Browse files
README.md
CHANGED
@@ -60,10 +60,11 @@ To simplify the comparison, we chosed the Pass@1 metric for the Python language,
|
|
60 |
|
61 |
| Model | HumanEval python pass@1 |
|
62 |
| --- |----------------------------------------------------------------------------- |
|
63 |
-
| phi-2 | 48.2% |
|
64 |
-
| **opencsg-phi-2-v0.1** |**54.3%**|
|
65 |
| stable-coder-3b | 29.3%|
|
66 |
| **opencsg-stable-coder-3b-v1**| **46.3%** |
|
|
|
|
|
|
|
67 |
|
68 |
|
69 |
|
@@ -162,10 +163,10 @@ HumanEval 是评估模型在代码生成方面性能的最常见的基准,尤
|
|
162 |
|
163 |
| 模型 | HumanEval python pass@1 |
|
164 |
| --- |----------------------------------------------------------------------------- |
|
165 |
-
| phi-2 | 48.2% |
|
166 |
-
| **opencsg-phi-2-v0.1** |**54.3%**|
|
167 |
| stable-coder-3b | 29.3%|
|
168 |
| **opencsg-stable-coder-3b-v1**| **46.3%** |
|
|
|
|
|
169 |
|
170 |
|
171 |
|
|
|
60 |
|
61 |
| Model | HumanEval python pass@1 |
|
62 |
| --- |----------------------------------------------------------------------------- |
|
|
|
|
|
63 |
| stable-coder-3b | 29.3%|
|
64 |
| **opencsg-stable-coder-3b-v1**| **46.3%** |
|
65 |
+
| phi-2 | 48.2% |
|
66 |
+
| **opencsg-phi-2-v0.1** |**54.3%**|
|
67 |
+
|
68 |
|
69 |
|
70 |
|
|
|
163 |
|
164 |
| 模型 | HumanEval python pass@1 |
|
165 |
| --- |----------------------------------------------------------------------------- |
|
|
|
|
|
166 |
| stable-coder-3b | 29.3%|
|
167 |
| **opencsg-stable-coder-3b-v1**| **46.3%** |
|
168 |
+
| phi-2 | 48.2% |
|
169 |
+
| **opencsg-phi-2-v0.1** |**54.3%**|
|
170 |
|
171 |
|
172 |
|