mpt-7b-instruct-sharded / generation_config.json
João Rafael
Replace model with mpt-7b-instruct, loaded in f16 and sharded to 2GB chunks
8d8911a
raw
history blame contribute delete
91 Bytes
{
"_from_model_config": true,
"transformers_version": "4.29.2",
"use_cache": false
}