Add pipeline tag and link to OpenCodeReasoning paper
Browse filesThis PR adds the `pipeline_tag` to the metadata to increase discoverability. It also adds a direct link to the Hugging Face paper page for the model and states that this model is described in the OpenCodeReasoning paper.
README.md
CHANGED
@@ -1,22 +1,23 @@
|
|
1 |
---
|
2 |
-
license: llama3.1
|
3 |
base_model:
|
4 |
- meta-llama/Llama-3.1-8B
|
5 |
datasets:
|
6 |
- nvidia/OpenMathInstruct-2
|
7 |
language:
|
8 |
- en
|
|
|
|
|
|
|
9 |
tags:
|
10 |
- nvidia
|
11 |
- math
|
12 |
-
library_name: transformers
|
13 |
---
|
14 |
|
15 |
# OpenMath2-Llama3.1-8B
|
16 |
|
17 |
-
OpenMath2-Llama3.1-8B is obtained by finetuning [Llama3.1-8B-Base](https://huggingface.co/meta-llama/Llama-3.1-8B) with [OpenMathInstruct-2](https://huggingface.co/datasets/nvidia/OpenMathInstruct-2).
|
18 |
|
19 |
-
The model outperforms [Llama3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on all the popular math benchmarks we evaluate on, especially on [MATH](https://github.com/hendrycks/math) by 15.9%.
|
20 |
|
21 |
<!-- <p align="center">
|
22 |
<img src="scaling_plot.jpg" width="350"><img src="math_level_comp.jpg" width="350">
|
@@ -81,7 +82,9 @@ pipeline = transformers.pipeline(
|
|
81 |
messages = [
|
82 |
{
|
83 |
"role": "user",
|
84 |
-
"content": "Solve the following math problem. Make sure to put the answer (and only answer) inside \\boxed{}
|
|
|
|
|
85 |
"What is the minimum value of $a^2+6a-7$?"},
|
86 |
]
|
87 |
|
|
|
1 |
---
|
|
|
2 |
base_model:
|
3 |
- meta-llama/Llama-3.1-8B
|
4 |
datasets:
|
5 |
- nvidia/OpenMathInstruct-2
|
6 |
language:
|
7 |
- en
|
8 |
+
library_name: transformers
|
9 |
+
license: llama3.1
|
10 |
+
pipeline_tag: text-generation
|
11 |
tags:
|
12 |
- nvidia
|
13 |
- math
|
|
|
14 |
---
|
15 |
|
16 |
# OpenMath2-Llama3.1-8B
|
17 |
|
18 |
+
OpenMath2-Llama3.1-8B is obtained by finetuning [Llama3.1-8B-Base](https://huggingface.co/meta-llama/Llama-3.1-8B) with [OpenMathInstruct-2](https://huggingface.co/datasets/nvidia/OpenMathInstruct-2). This model is described in the paper [OpenCodeReasoning: Advancing Data Distillation for Competitive Coding](https://huggingface.co/papers/2504.01943).
|
19 |
|
20 |
+
The model outperforms [Llama3.1-8B-Instruct](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct) on all the popular math benchmarks we evaluate on, especially on [MATH](https://github.com/hendrycks/math) by 15.9%.
|
21 |
|
22 |
<!-- <p align="center">
|
23 |
<img src="scaling_plot.jpg" width="350"><img src="math_level_comp.jpg" width="350">
|
|
|
82 |
messages = [
|
83 |
{
|
84 |
"role": "user",
|
85 |
+
"content": "Solve the following math problem. Make sure to put the answer (and only answer) inside \\boxed{}.
|
86 |
+
|
87 |
+
" +
|
88 |
"What is the minimum value of $a^2+6a-7$?"},
|
89 |
]
|
90 |
|