llamav2 / llama-2-7b-chat /params.json
vietanhdev's picture
Add 7B models
f16d19f
raw
history blame contribute delete
102 Bytes
{"dim": 4096, "multiple_of": 256, "n_heads": 32, "n_layers": 32, "norm_eps": 1e-06, "vocab_size": -1}