๋ชจ๋ธ : llama3 8B

๋ฐ์ดํ„ฐ : AIhub ์š”์•ฝ ๋ฐ์ดํ„ฐ์…‹, ํ—ˆ๊น…ํŽ˜์ด์Šค (MIT, Apache ๋ฐ์ดํ„ฐ์…‹ ์ผ๋ถ€)

๋ชจ๋ธ ํ•™์Šต ํŒŒ๋ผ๋ฏธํ„ฐ : ๋ชจ๋ธ์˜ ํŒŒ๋ผ๋ฏธํ„ฐ ์ค‘ 10%์ •๋„๋ฅผ trainableํ•œ parameter๋กœ ์„ค์ •

generate config :

max_new_tokens = 1024,

repetition_penalty = 1.1, #1.1~1.2

temperature = 0.75,

do_sample= True

top_k = 5

์ฃผ์˜ : ๋ฏผ๊ฐ ๋ฐ์ดํ„ฐ๊ฐ€ ๋“ค์–ด๊ฐ€ ์žˆ์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

Downloads last month
2,259
Safetensors
Model size
8.03B params
Tensor type
FP16
ยท
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and the model is not deployed on the HF Inference API.

Model tree for sel303/llama3-diverce-ver1.0

Quantizations
2 models

Spaces using sel303/llama3-diverce-ver1.0 6