You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have already tried simple summation and averaging methods, as well as weighted averaging based on the data volume of each sub-dataset, but none of these methods can reproduce the 'Avg.' value in the paper. How exactly should this value be calculated? Also, is the ToolGen model related to In-domain not yet open-sourced?
The text was updated successfully, but these errors were encountered:
Hi, it is weighted average. The sample number for I1, I2, and I3 Instruction is 163, 106, 61 respectively and for I1 Tool, I1 Cat, I2 Cat is 153, 158, 124 respectively.
Also, is the ToolGen model related to In-domain not yet open-sourced?
Sorry that they are actually open-sourced, but I forgot to add them to the collection, they are now at https://huggingface.co/collections/reasonwang/toolgen-668a46a4959745ec8e9891f6, ending with G1, G2, and G3.
I have already tried simple summation and averaging methods, as well as weighted averaging based on the data volume of each sub-dataset, but none of these methods can reproduce the 'Avg.' value in the paper. How exactly should this value be calculated? Also, is the ToolGen model related to In-domain not yet open-sourced?
The text was updated successfully, but these errors were encountered: