compasszzn commited on
Commit
164b617
β€’
1 Parent(s): 2f23fe4

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +32 -0
README.md CHANGED
@@ -14,6 +14,15 @@ language:
14
  - fr
15
  - bn
16
  ---
 
 
 
 
 
 
 
 
 
17
  ### Introduction
18
 
19
  We introduce πŸ™ MathOctopus, a series of open-source large language models (LLMs) specifically tailored for multilingual math problem-solving. The MathOctopus models are trained on πŸ€— MGSM8KInstruct Dataset, encompassing ten distinct languages.
@@ -61,7 +70,9 @@ Our dataset and models are all available at Huggingface.
61
  *-Cross refers to our model trained with cross-training strategy.
62
 
63
  *-xRFT means we train the model with multilingual rejection sampling.
 
64
  ### **Overall Results on MGSM**
 
65
  | 7B Model | En | Sw | Zh | Bn | De | Es | Fr | Ja | Ru | Th | Overall |
66
  |:--------------------------------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|
67
  | MathOctopus<sup>C</sup> | 52.0 | 23.6 | 31.6 | 18.8 | 38.0 | 39.2 | 36.4 | 27.2 | 33.6 | 21.6 | 32.2 |
@@ -85,7 +96,10 @@ Our dataset and models are all available at Huggingface.
85
  | **xRFT**-MathOctopus<sup>C</sup>| 53.6 | 27.6 | 34.4 | 19.2 | 47.2 | 47.6 | 44.8 | 30.8 | 38.8 | 22.8 | 36.7 |
86
  | MathOctopus<sup>P</sup> | 56.4 | 46.8 | 52.0 | 35.2 | 47.2 | 53.2 | 48.0 | 39.2 | 45.6 | 41.2 | 46.5 |
87
  | **xRFT**-MathOctopus<sup>P</sup>| 51.6 | 47.2 | 52.4 | 37.6 | 51.2 | 52.8 | 44.4 | 41.6 | 50.0 | 47.6 | 47.6 |
 
 
88
  ### **Overall Results on MSVAMP**
 
89
  | 7B Model | En | Sw | Zh | Bn | De | Es | Fr | Ja | Ru | Th | Overall |
90
  |:--------------------------------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|
91
  | MathOctopus<sup>C</sup> | 49.2 | 36.6 | 43.6 | 30.2 | 48.6 | 46.8 | 46.4 | 42.5 | 46.7 | 34.0 | 42.5 |
@@ -109,7 +123,10 @@ Our dataset and models are all available at Huggingface.
109
  | **xRFT**-MathOctopus<sup>C</sup>| 48.1 | 42.8 | 43.6 | 23.3 | 48.7 | 50.0 | 48.9 | 43.4 | 44.6 | 35.5 | 42.9 |
110
  | MathOctopus<sup>P</sup> | 56.4 | 46.8 | 52.0 | 35.2 | 47.2 | 53.2 | 48.0 | 39.2 | 45.6 | 41.2 | 46.5 |
111
  | **xRFT**-MathOctopus<sup>P</sup>| 48.0 | 42.3 | 46.1 | 36.2 | 47.5 | 48.5 | 48.3 | 45.8 | 47.2 | 41.2 | 45.1 |
 
 
112
  ### **MathOctopus in English**
 
113
  | Models | GSM8K | SVAMP |
114
  |:--------------------------------|:--------|:--------|
115
  | LLaMA 2-7B | 42.4 | 38.3 |
@@ -121,5 +138,20 @@ Our dataset and models are all available at Huggingface.
121
  | LLaMA 1-33B | 50.0 | 49.0 |
122
  | MathOctopus<sup>P</sup>-33B | 56.0 | 52.5 |
123
  | MathOctopus<sup>C</sup>-33B | 53.7 | 51.5 |
 
124
  ## Intended Uses
125
  These models are trained for research purposes. They are designed to solve multilingual math problems. They can be used in educational software, tutoring systems, or any application where a solution to a math problem is needed.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14
  - fr
15
  - bn
16
  ---
17
+
18
+ # πŸ™ Breaking Language Barriers in Multilingual Mathematical Reasoning: Insights and Observations
19
+
20
+ Project Page: [https://mathoctopus.github.io/](https://mathoctopus.github.io/)
21
+
22
+ Paper: [https://arxiv.org/abs/2310.20246.pdf](https://arxiv.org/abs/2310.20246.pdf)
23
+
24
+ Code: [https://github.com/microsoft/MathOctopus](https://github.com/microsoft/MathOctopus)
25
+
26
  ### Introduction
27
 
28
  We introduce πŸ™ MathOctopus, a series of open-source large language models (LLMs) specifically tailored for multilingual math problem-solving. The MathOctopus models are trained on πŸ€— MGSM8KInstruct Dataset, encompassing ten distinct languages.
 
70
  *-Cross refers to our model trained with cross-training strategy.
71
 
72
  *-xRFT means we train the model with multilingual rejection sampling.
73
+
74
  ### **Overall Results on MGSM**
75
+
76
  | 7B Model | En | Sw | Zh | Bn | De | Es | Fr | Ja | Ru | Th | Overall |
77
  |:--------------------------------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|
78
  | MathOctopus<sup>C</sup> | 52.0 | 23.6 | 31.6 | 18.8 | 38.0 | 39.2 | 36.4 | 27.2 | 33.6 | 21.6 | 32.2 |
 
96
  | **xRFT**-MathOctopus<sup>C</sup>| 53.6 | 27.6 | 34.4 | 19.2 | 47.2 | 47.6 | 44.8 | 30.8 | 38.8 | 22.8 | 36.7 |
97
  | MathOctopus<sup>P</sup> | 56.4 | 46.8 | 52.0 | 35.2 | 47.2 | 53.2 | 48.0 | 39.2 | 45.6 | 41.2 | 46.5 |
98
  | **xRFT**-MathOctopus<sup>P</sup>| 51.6 | 47.2 | 52.4 | 37.6 | 51.2 | 52.8 | 44.4 | 41.6 | 50.0 | 47.6 | 47.6 |
99
+
100
+
101
  ### **Overall Results on MSVAMP**
102
+
103
  | 7B Model | En | Sw | Zh | Bn | De | Es | Fr | Ja | Ru | Th | Overall |
104
  |:--------------------------------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|:--------|
105
  | MathOctopus<sup>C</sup> | 49.2 | 36.6 | 43.6 | 30.2 | 48.6 | 46.8 | 46.4 | 42.5 | 46.7 | 34.0 | 42.5 |
 
123
  | **xRFT**-MathOctopus<sup>C</sup>| 48.1 | 42.8 | 43.6 | 23.3 | 48.7 | 50.0 | 48.9 | 43.4 | 44.6 | 35.5 | 42.9 |
124
  | MathOctopus<sup>P</sup> | 56.4 | 46.8 | 52.0 | 35.2 | 47.2 | 53.2 | 48.0 | 39.2 | 45.6 | 41.2 | 46.5 |
125
  | **xRFT**-MathOctopus<sup>P</sup>| 48.0 | 42.3 | 46.1 | 36.2 | 47.5 | 48.5 | 48.3 | 45.8 | 47.2 | 41.2 | 45.1 |
126
+
127
+
128
  ### **MathOctopus in English**
129
+
130
  | Models | GSM8K | SVAMP |
131
  |:--------------------------------|:--------|:--------|
132
  | LLaMA 2-7B | 42.4 | 38.3 |
 
138
  | LLaMA 1-33B | 50.0 | 49.0 |
139
  | MathOctopus<sup>P</sup>-33B | 56.0 | 52.5 |
140
  | MathOctopus<sup>C</sup>-33B | 53.7 | 51.5 |
141
+
142
  ## Intended Uses
143
  These models are trained for research purposes. They are designed to solve multilingual math problems. They can be used in educational software, tutoring systems, or any application where a solution to a math problem is needed.
144
+
145
+ ## Citation
146
+ Please cite our paper if you use our data, model or code. Please also kindly cite the original dataset papers.
147
+
148
+ ```
149
+ @misc{chen2023breaking,
150
+ title={Breaking Language Barriers in Multilingual Mathematical Reasoning: Insights and Observations},
151
+ author={Nuo Chen and Zinan Zheng and Ning Wu and Linjun Shou and Ming Gong and Yangqiu Song and Dongmei Zhang and Jia Li},
152
+ year={2023},
153
+ eprint={2310.20246},
154
+ archivePrefix={arXiv},
155
+ primaryClass={cs.CL}
156
+ }
157
+ ```