huggingface/candle
1
0
Fork 0
mirror of https://github.com/huggingface/candle.git synced 2025-06-16 18:48:51 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: add missingtop_p in llama_multiprocess (#905)

Browse Source
This commit is contained in:
Mahmoud
2023-09-20 00:54:56 -07:00
committed by GitHub
parent 05626ef492
commit 098dd0d1e9
1 changed files with 5 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
candle-examples/examples/llama_multiprocess/main.rs
View File

@ -89,6 +89,10 @@ struct Args {
#[arg(long)] #[arg(long)]
temperature: Option<f64>, temperature: Option<f64>,
/// Nucleus sampling probability cutoff.
#[arg(long)]
top_p: Option<f64>,
/// The seed to use when generating random samples. /// The seed to use when generating random samples.
#[arg(long, default_value_t = 299792458)] #[arg(long, default_value_t = 299792458)]
seed: u64, seed: u64,
@ -222,7 +226,7 @@ fn main() -> Result<()> {
.to_vec(); .to_vec();
println!("starting the inference loop"); println!("starting the inference loop");
let mut logits_processor = LogitsProcessor::new(args.seed, args.temperature); let mut logits_processor = LogitsProcessor::new(args.seed, args.temperature, args.top_p);
let mut new_tokens = vec![]; let mut new_tokens = vec![];
let start_gen = std::time::Instant::now(); let start_gen = std::time::Instant::now();
let mut index_pos = 0; let mut index_pos = 0;