Update README.md
Browse files
README.md
CHANGED
@@ -20,7 +20,7 @@ model-index:
|
|
20 |
|
21 |
We used LoRa to further pre-train Meta's CodeLLaMA-7B-hf model with high-quality C++ code tokens.
|
22 |
|
23 |
-
Furthermore, we
|
24 |
|
25 |
## Model Details
|
26 |
|
@@ -35,10 +35,7 @@ We pre-trained CodeLLaMA-7B further using 543 GB of C++ code collected online, a
|
|
35 |
## Requirements
|
36 |
|
37 |
```
|
38 |
-
|
39 |
-
tokenizers==0.13.3
|
40 |
-
transformers==4.33.0
|
41 |
-
bitsandbytes==0.41.1
|
42 |
```
|
43 |
|
44 |
## How to reproduce HumanEval-X results
|
@@ -86,7 +83,7 @@ pipeline = transformers.pipeline(
|
|
86 |
)
|
87 |
|
88 |
sequences = pipeline(
|
89 |
-
'
|
90 |
do_sample=True,
|
91 |
top_k=10,
|
92 |
temperature=0.1,
|
|
|
20 |
|
21 |
We used LoRa to further pre-train Meta's CodeLLaMA-7B-hf model with high-quality C++ code tokens.
|
22 |
|
23 |
+
Furthermore, we fine-tuned on CodeM's C++ instruction data.
|
24 |
|
25 |
## Model Details
|
26 |
|
|
|
35 |
## Requirements
|
36 |
|
37 |
```
|
38 |
+
pip install torch transformers accelerate
|
|
|
|
|
|
|
39 |
```
|
40 |
|
41 |
## How to reproduce HumanEval-X results
|
|
|
83 |
)
|
84 |
|
85 |
sequences = pipeline(
|
86 |
+
'#include <iostream>\n#include <vector>\n\nusing namespace std;\n\nvoid quickSort(int *data, int start, int end) {',
|
87 |
do_sample=True,
|
88 |
top_k=10,
|
89 |
temperature=0.1,
|