diff options
| author | Georgi Gerganov <ggerganov@gmail.com> | 2024-02-25 12:09:09 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-02-25 12:09:09 +0200 |
| commit | ab336a9d5e5352ecdcdf4c12d2d54cf4ef82ce31 (patch) | |
| tree | 5694ecb0647b10a6377a273737b63bb025dc961d /common/common.h | |
| parent | 69917dfa55674c608360638bb4d6a12a315e2810 (diff) | |
code : normalize enum names (#5697)
* coda : normalize enum names
ggml-ci
* code : cont
* code : cont
Diffstat (limited to 'common/common.h')
| -rw-r--r-- | common/common.h | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/common/common.h b/common/common.h index 935771d4..3e21579b 100644 --- a/common/common.h +++ b/common/common.h @@ -61,7 +61,7 @@ struct gpt_params { float p_split = 0.1f; // speculative decoding split probability int32_t n_gpu_layers = -1; // number of layers to store in VRAM (-1 - use default) int32_t n_gpu_layers_draft = -1; // number of layers to store in VRAM for the draft model (-1 - use default) - llama_split_mode split_mode = LLAMA_SPLIT_LAYER; // how to split the model across GPUs + llama_split_mode split_mode = LLAMA_SPLIT_MODE_LAYER; // how to split the model across GPUs int32_t main_gpu = 0; // the GPU that is used for scratch and small tensors float tensor_split[128] = {0}; // how split tensors should be distributed across GPUs int32_t n_beams = 0; // if non-zero then use beam search of given width. @@ -75,7 +75,7 @@ struct gpt_params { float yarn_beta_fast = 32.0f; // YaRN low correction dim float yarn_beta_slow = 1.0f; // YaRN high correction dim int32_t yarn_orig_ctx = 0; // YaRN original context length - int32_t rope_scaling_type = LLAMA_ROPE_SCALING_UNSPECIFIED; + int32_t rope_scaling_type = LLAMA_ROPE_SCALING_TYPE_UNSPECIFIED; ggml_numa_strategy numa = GGML_NUMA_STRATEGY_DISABLED; // // sampling parameters |