PAL_llama-3-8B / generation_config.json
anthony-lemurian's picture
This is llama 3 8B quantized (weights only) into PAL4 (2, 1.278316448612589) with blocksize 16.
0d27034 verified
raw
history blame
172 Bytes
{
"bos_token_id": 128000,
"do_sample": true,
"eos_token_id": 128001,
"max_length": 4096,
"temperature": 0.6,
"top_p": 0.9,
"transformers_version": "4.41.2"
}