extern/whisper.cpp
1
0
Fork 1
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2025-08-09 19:28:39 +02:00
Code Issues Packages Projects Releases Wiki Activity

talk-llama : sync llama.cpp

Browse Source
This commit is contained in:
Georgi Gerganov
2024-02-25 20:00:10 +02:00
parent 7a6e385c1b
commit 3170841ed9
2 changed files with 89 additions and 61 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
examples/talk-llama/llama.h
View File

@ -102,28 +102,30 @@ extern "C" {
LLAMA_FTYPE_MOSTLY_IQ3_XXS = 23, // except 1d tensors
LLAMA_FTYPE_MOSTLY_IQ1_S = 24, // except 1d tensors
LLAMA_FTYPE_MOSTLY_IQ4_NL = 25, // except 1d tensors
LLAMA_FTYPE_MOSTLY_IQ3_S = 26, // except 1d tensors
LLAMA_FTYPE_MOSTLY_IQ3_M = 27, // except 1d tensors
LLAMA_FTYPE_GUESSED = 1024, // not specified in the model file
};
enum llama_rope_scaling_type {
LLAMA_ROPE_SCALING_UNSPECIFIED = -1,
LLAMA_ROPE_SCALING_NONE = 0,
LLAMA_ROPE_SCALING_LINEAR = 1,
LLAMA_ROPE_SCALING_YARN = 2,
LLAMA_ROPE_SCALING_MAX_VALUE = LLAMA_ROPE_SCALING_YARN,
LLAMA_ROPE_SCALING_TYPE_UNSPECIFIED = -1,
LLAMA_ROPE_SCALING_TYPE_NONE = 0,
LLAMA_ROPE_SCALING_TYPE_LINEAR = 1,
LLAMA_ROPE_SCALING_TYPE_YARN = 2,
LLAMA_ROPE_SCALING_TYPE_MAX_VALUE = LLAMA_ROPE_SCALING_TYPE_YARN,
};
enum llama_pooling_type {
LLAMA_POOLING_NONE = 0,
LLAMA_POOLING_MEAN = 1,
LLAMA_POOLING_CLS = 2,
LLAMA_POOLING_TYPE_NONE = 0,
LLAMA_POOLING_TYPE_MEAN = 1,
LLAMA_POOLING_TYPE_CLS = 2,
};
enum llama_split_mode {
LLAMA_SPLIT_NONE = 0, // single GPU
LLAMA_SPLIT_LAYER = 1, // split layers and KV across GPUs
LLAMA_SPLIT_ROW = 2, // split rows across GPUs
LLAMA_SPLIT_MODE_NONE = 0, // single GPU
LLAMA_SPLIT_MODE_LAYER = 1, // split layers and KV across GPUs
LLAMA_SPLIT_MODE_ROW = 2, // split rows across GPUs
};
typedef struct llama_token_data {
@ -171,9 +173,9 @@ extern "C" {
} llama_batch;
enum llama_model_kv_override_type {
LLAMA_KV_OVERRIDE_INT,
LLAMA_KV_OVERRIDE_FLOAT,
LLAMA_KV_OVERRIDE_BOOL,
LLAMA_KV_OVERRIDE_TYPE_INT,
LLAMA_KV_OVERRIDE_TYPE_FLOAT,
LLAMA_KV_OVERRIDE_TYPE_BOOL,
};
struct llama_model_kv_override {