google-deepmind/gemma-2b-it

2B instruct version of Google’s Gemma model

Capabilities

Max TokensTop-P

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`max_new_tokens`	integer	Maximum number of tokens to generate. A word is generally 2-3 tokens	`200`	min: 1
`min_new_tokens`	integer	Minimum number of tokens to generate. To disable, set to -1. A word is generally 2-3 tokens.	`-1`	min: -1
`prompt`	string	Prompt to send to the model.	`"Write me a poem about Machine Learning."`	—
`repetition_penalty`	number	A parameter that controls how repetitive text can be. Lower means more repetitive, while higher means less repetitive. Set to 1.0 to disable.	`1.15`	min: 0
`temperature`	number	Adjusts randomness of outputs, greater than 1 is random and 0 is deterministic, 0.75 is a good starting value.	`0.7`	min: 0.01, max: 5
`top_k`	integer	When decoding text, samples from the top k most likely tokens; lower to ignore less likely tokens	`50`	—
`top_p`	number	When decoding text, samples from the top p percentage of most likely tokens; lower to ignore less likely tokens	`0.95`	min: 0, max: 1

max_new_tokensinteger

Maximum number of tokens to generate. A word is generally 2-3 tokens

Default: 200min: 1

min_new_tokensinteger

Minimum number of tokens to generate. To disable, set to -1. A word is generally 2-3 tokens.

Default: -1min: -1

promptstring

Prompt to send to the model.

Default: "Write me a poem about Machine Learning."

repetition_penaltynumber

A parameter that controls how repetitive text can be. Lower means more repetitive, while higher means less repetitive. Set to 1.0 to disable.

Default: 1.15min: 0

temperaturenumber

Adjusts randomness of outputs, greater than 1 is random and 0 is deterministic, 0.75 is a good starting value.

Default: 0.7min: 0.01, max: 5

top_kinteger

When decoding text, samples from the top k most likely tokens; lower to ignore less likely tokens

Default: 50

top_pnumber

When decoding text, samples from the top p percentage of most likely tokens; lower to ignore less likely tokens

Default: 0.95min: 0, max: 1

Version: dff94eaf770eUpdated: 7/25/2026134.4K runs