Skip to content

Commit

Permalink
finetune : add --n-gpu-layers flag info to --help (ggerganov#4128)
Browse files Browse the repository at this point in the history
  • Loading branch information
csaben authored and olexiyb committed Nov 23, 2023
1 parent 357923e commit 113e1a4
Showing 1 changed file with 1 addition and 0 deletions.
1 change: 1 addition & 0 deletions examples/finetune/finetune.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -1288,6 +1288,7 @@ static void train_print_usage(int argc, char ** argv, const struct train_params
fprintf(stderr, " --model-base FNAME model path from which to load base model (default '%s')\n", params->fn_model_base);
fprintf(stderr, " --lora-out FNAME path to save llama lora (default '%s')\n", params->fn_lora_out);
fprintf(stderr, " --only-write-lora only save llama lora, don't do any training. use this if you only want to convert a checkpoint to a lora adapter.\n");
fprintf(stderr, " --n-gpu-layers N Number of model layers to offload to GPU (default 0).\n");
fprintf(stderr, " --norm-rms-eps F RMS-Norm epsilon value (default %f)\n", params->f_norm_rms_eps);
fprintf(stderr, " --rope-freq-base F Frequency base for ROPE (default %f)\n", params->rope_freq_base);
fprintf(stderr, " --rope-freq-scale F Frequency scale for ROPE (default %f)\n", params->rope_freq_scale);
Expand Down

0 comments on commit 113e1a4

Please sign in to comment.