Update README.md

README.md

@@ -18,13 +18,6 @@ This repository contains:
 
 Note: Our models are tuned based on Codellama, thus, licenses applicable to Codellama, such as [Llama license](https://ai.meta.com/resources/models-and-libraries/llama-downloads/), also hold on these models
 
-|                                                                    |  Top-1 | Voting@100 | Rerank@100 |
-|--------------------------------------------------------------------|:------:|:----------:|:----------:|
-| Codellama-7b-hf-SFT-warmup-GSM8k                                   |  63.00 |      -     |      -     |
-| Codellama-7b-hf-SFT-GSM8k<br>(+Codellama-7b-hf-SFT-Rerank-GSM8k)   | 63.68  |    68.0    |    77.0    |
-| Codellama-7b-hf-ReFT-GSM8k<br>(+Codellama-7b-hf-ReFT-Rerank-GSM8k) | 75.28  |    78.0    |    81.2    |
-
-
 ## Training Data
 The model is trained on GSM8k data with Python SDP CoT format, which can be found [here](https://github.com/lqtrung1998/mwp_ReFT)
 
@@ -38,6 +31,14 @@ Rerank model is trained to classify if the output CoT is correct or not using sa
 ## Evaluation Results
 See evaluations results of the models at table 4 of the research paper.
 
+Updated results:
+|                                                                    |  Top-1 | Voting@100 | Rerank@100 |
+|--------------------------------------------------------------------|:------:|:----------:|:----------:|
+| Codellama-7b-hf-SFT-warmup-GSM8k                                   |  63.00 |      -     |      -     |
+| Codellama-7b-hf-SFT-GSM8k<br>(+Codellama-7b-hf-SFT-Rerank-GSM8k)   | 63.68  |    68.0    |    77.0    |
+| Codellama-7b-hf-ReFT-GSM8k<br>(+Codellama-7b-hf-ReFT-Rerank-GSM8k) | 75.28  |    78.0    |    81.2    |
+
+
 ## Usage
 You can use the models through Huggingface's Transformers library or follow scripts in our repo.