Hugging Face's logo

codefuse-ai
/
CodeFuse-CodeLlama-34B

Text Generation
Transformers
PyTorch
llama
text-generation-inference
Inference Endpoints
Model card Files Community
10
codefuse-admin commited on
Commit
7b92b5b
•
1 Parent(s): fbd6063

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +26 -10
README.md CHANGED
@@ -26,11 +26,19 @@ The context length of finetuning is 4K while it is able to be finetuned by 16k c
26
 
27
  ## Performance
28
 
29
- | Model | HumanEval(pass@1) |
30
- | :---------------------------- | :---------------: |
31
- | CodeLlama-34b | 48.8%(greedy decoding) |
32
- | CodeLlama-34b-Python | 53.7%(greedy decoding) |
33
- | **CodeFuse-CodeLlama-34B** | **74.4%**(greedy decoding) |
 
 
 
 
 
 
 
 
34
 
35
  <br>
36
 
@@ -130,11 +138,19 @@ CodeFuse-CodeLlama34B-MFT 是一个通过QLoRA对基座模型CodeLlama-34b-Pytho
130
 
131
  ## 评测表现(代码)
132
 
133
- | 模型 | HumanEval(pass@1) |
134
- | :---------------------------- | :---------------: |
135
- | CodeLlama-34b | 48.8%(greedy decoding) |
136
- | CodeLlama-34b-Python | 53.7%(greedy decoding) |
137
- | **CodeFuse-CodeLlama-34B** | **74.4%**(greedy decoding) |
 
 
 
 
 
 
 
 
138
  <br>
139
 
140
  ## Requirements
 
26
 
27
  ## Performance
28
 
29
+
30
+ | Model | HumanEval(pass@1) | Date |
31
+ |:----------------------------|:-----------------:|:-------:|
32
+ | **CodeFuse-CodeLlama-34B** | **74.4%** | 2023.9 |
33
+ | WizardCoder-Python-34B-V1.0 | 73.2% | 2023.8 |
34
+ | GPT-4(zero-shot) | 67.0% | 2023.3 |
35
+ | PanGu-Coder2 15B | 61.6% | 2023.8 |
36
+ | CodeLlama-34b-Python | 53.7% | 2023.8 |
37
+ | CodeLlama-34b | 48.8% | 2023.8 |
38
+ | GPT-3.5(zero-shot) | 48.1% | 2022.11 |
39
+ | OctoCoder | 46.2% | 2023.8 |
40
+ | StarCoder-15B | 33.6% | 2023.5 |
41
+ | LLaMA 2 70B(zero-shot) | 29.9% | 2023.7 |
42
 
43
  <br>
44
 
 
138
 
139
  ## 评测表现(代码)
140
 
141
+ | 模型 | HumanEval(pass@1) | 日期 |
142
+ |:----------------------------|:-----------------:|:-------:|
143
+ | **CodeFuse-CodeLlama-34B** | **74.4%** | 2023.9 |
144
+ | WizardCoder-Python-34B-V1.0 | 73.2% | 2023.8 |
145
+ | GPT-4(zero-shot) | 67.0% | 2023.3 |
146
+ | PanGu-Coder2 15B | 61.6% | 2023.8 |
147
+ | CodeLlama-34b-Python | 53.7% | 2023.8 |
148
+ | CodeLlama-34b | 48.8% | 2023.8 |
149
+ | GPT-3.5(zero-shot) | 48.1% | 2022.11 |
150
+ | OctoCoder | 46.2% | 2023.8 |
151
+ | StarCoder-15B | 33.6% | 2023.5 |
152
+ | LLaMA 2 70B(zero-shot) | 29.9% | 2023.7 |
153
+
154
  <br>
155
 
156
  ## Requirements