evalplus / evalplus

Rigourous evaluation of LLM-synthesized code - NeurIPS 2023 & COLM 2024
https://evalplus.github.io
Apache License 2.0
1.22k stars 107 forks source link

google/gemma-7b-it #107

Closed msoftware closed 6 months ago

msoftware commented 8 months ago

Model introduction

This model card corresponds to the 7B instruct version of the Gemma model. You can also visit the model card of the 2B base model, 7B base model, and 2B instruct model.

Model URL

https://huggingface.co/google/gemma-7b-it

Data type

float16

Additional instructions (Optional)

No response

Author

No

Security

Integrity

FatPigeorz commented 6 months ago

gemma-2b-it

HE+:

image

MBPP+:

image
FatPigeorz commented 6 months ago

gemma-2b

HE+:

image

MBPP+

image
FatPigeorz commented 6 months ago

gemma-7b-it

HE+:

image

MBPP+:

image
FatPigeorz commented 6 months ago

gemma-7b

HE+:

image

MBPP+:

image