vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama : switch to floating-point token positions

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2024-02-23 12:18:30 +02:00
parent 15499eb942
commit fc775366f1
No known key found for this signature in database
GPG key ID: BF970631944C16B7
14 changed files with 68 additions and 61 deletions
Download patch file Download diff file Expand all files Collapse all files

8
ggml-metal.m
View file

@ -2057,7 +2057,13 @@ static bool ggml_metal_graph_compute(
// skip 3, n_ctx, used in GLM RoPE, unimplemented in metal
const int n_orig_ctx = ((int32_t *) dst->op_params)[4];
float freq_base, freq_scale, ext_factor, attn_factor, beta_fast, beta_slow;
float freq_base;
float freq_scale;
float ext_factor;
float attn_factor;
float beta_fast;
float beta_slow;
memcpy(&freq_base, (int32_t *) dst->op_params + 5, sizeof(float));
memcpy(&freq_scale, (int32_t *) dst->op_params + 6, sizeof(float));
memcpy(&ext_factor, (int32_t *) dst->op_params + 7, sizeof(float));