Questions about data scale
#9
by
masterLan
- opened
How much data was used to train the final version of Qwen-2.5-MATH-PRM?
While reading the paper, I noticed that in Figure 8, the score for ProcessBench is 66.5. However, in the final results presented in Table 7, the score for Qwen-2.5-MATH-PRM is 73.5. Additionally, the data in Figure 8 does not match the data in Table 4, which raises some concerns.
