root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-25 10:54:36 +00:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

llama : streamline embeddings from "non-embedding" models (#8087)

Browse Source
This commit is contained in:
Douglas Hanley 2024-07-05 02:05:56 -05:00 committed by GitHub
parent bcefa03bc0
commit d12f781074
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 36 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
common/common.cpp
View File

@ -472,6 +472,14 @@ bool gpt_params_find_arg(int argc, char ** argv, const std::string & arg, gpt_pa
else { invalid_param = true; } else { invalid_param = true; }
return true; return true;
} }
if (arg == "--attention") {
CHECK_ARG
std::string value(argv[i]);
/**/ if (value == "causal") { params.attention_type = LLAMA_ATTENTION_TYPE_CAUSAL; }
else if (value == "non-causal") { params.attention_type = LLAMA_ATTENTION_TYPE_NON_CAUSAL; }
else { invalid_param = true; }
return true;
}
if (arg == "--defrag-thold" || arg == "-dt") { if (arg == "--defrag-thold" || arg == "-dt") {
CHECK_ARG CHECK_ARG
params.defrag_thold = std::stof(argv[i]); params.defrag_thold = std::stof(argv[i]);
@ -1468,8 +1476,10 @@ void gpt_params_print_usage(int /*argc*/, char ** argv, const gpt_params & param
"For schemas w/ external $refs, use --grammar + example/json_schema_to_grammar.py instead" }); "For schemas w/ external $refs, use --grammar + example/json_schema_to_grammar.py instead" });
options.push_back({ "embedding" }); options.push_back({ "embedding" });
options.push_back({ "embedding", " --pooling {none,mean,cls}", options.push_back({ "embedding", " --pooling {none,mean,cls,last}",
"pooling type for embeddings, use model default if unspecified" }); "pooling type for embeddings, use model default if unspecified" });
options.push_back({ "embedding", " --attention {causal,non-causal}",
"attention type for embeddings, use model default if unspecified" });
options.push_back({ "context hacking" }); options.push_back({ "context hacking" });
options.push_back({ "*", " --rope-scaling {none,linear,yarn}", options.push_back({ "*", " --rope-scaling {none,linear,yarn}",
@ -2175,6 +2185,7 @@ struct llama_context_params llama_context_params_from_gpt_params(const gpt_param
cparams.yarn_beta_slow = params.yarn_beta_slow; cparams.yarn_beta_slow = params.yarn_beta_slow;
cparams.yarn_orig_ctx = params.yarn_orig_ctx; cparams.yarn_orig_ctx = params.yarn_orig_ctx;
cparams.pooling_type = params.pooling_type; cparams.pooling_type = params.pooling_type;
cparams.attention_type = params.attention_type;
cparams.defrag_thold = params.defrag_thold; cparams.defrag_thold = params.defrag_thold;
cparams.cb_eval = params.cb_eval; cparams.cb_eval = params.cb_eval;
cparams.cb_eval_user_data = params.cb_eval_user_data; cparams.cb_eval_user_data = params.cb_eval_user_data;

1
common/common.h
View File

@ -99,6 +99,7 @@ struct gpt_params {
enum llama_split_mode split_mode = LLAMA_SPLIT_MODE_LAYER; // how to split the model across GPUs enum llama_split_mode split_mode = LLAMA_SPLIT_MODE_LAYER; // how to split the model across GPUs
enum llama_rope_scaling_type rope_scaling_type = LLAMA_ROPE_SCALING_TYPE_UNSPECIFIED; enum llama_rope_scaling_type rope_scaling_type = LLAMA_ROPE_SCALING_TYPE_UNSPECIFIED;
enum llama_pooling_type pooling_type = LLAMA_POOLING_TYPE_UNSPECIFIED; // pooling type for embeddings enum llama_pooling_type pooling_type = LLAMA_POOLING_TYPE_UNSPECIFIED; // pooling type for embeddings
enum llama_attention_type attention_type = LLAMA_ATTENTION_TYPE_UNSPECIFIED; // attention type for embeddings
// // sampling parameters // // sampling parameters
struct llama_sampling_params sparams; struct llama_sampling_params sparams;

7
include/llama.h
View File

@ -180,6 +180,12 @@ extern "C" {
LLAMA_POOLING_TYPE_LAST = 3, LLAMA_POOLING_TYPE_LAST = 3,
}; };
enum llama_attention_type {
LLAMA_ATTENTION_TYPE_UNSPECIFIED = -1,
LLAMA_ATTENTION_TYPE_CAUSAL = 0,
LLAMA_ATTENTION_TYPE_NON_CAUSAL = 1,
};
enum llama_split_mode { enum llama_split_mode {
LLAMA_SPLIT_MODE_NONE = 0, // single GPU LLAMA_SPLIT_MODE_NONE = 0, // single GPU
LLAMA_SPLIT_MODE_LAYER = 1, // split layers and KV across GPUs LLAMA_SPLIT_MODE_LAYER = 1, // split layers and KV across GPUs
@ -297,6 +303,7 @@ extern "C" {
enum llama_rope_scaling_type rope_scaling_type; // RoPE scaling type, from `enum llama_rope_scaling_type` enum llama_rope_scaling_type rope_scaling_type; // RoPE scaling type, from `enum llama_rope_scaling_type`
enum llama_pooling_type pooling_type; // whether to pool (sum) embedding results by sequence id enum llama_pooling_type pooling_type; // whether to pool (sum) embedding results by sequence id
enum llama_attention_type attention_type; // attention type to use for embeddings
// ref: https://github.com/ggerganov/llama.cpp/pull/2054 // ref: https://github.com/ggerganov/llama.cpp/pull/2054
float rope_freq_base; // RoPE base frequency, 0 = from model float rope_freq_base; // RoPE base frequency, 0 = from model

25
src/llama.cpp
View File

@ -13840,7 +13840,7 @@ static void llama_set_inputs(llama_context & lctx, const llama_batch & batch) {
} }
} }
if (cparams.pooling_type == LLAMA_POOLING_TYPE_MEAN) { if (cparams.embeddings && cparams.pooling_type == LLAMA_POOLING_TYPE_MEAN) {
const int64_t n_tokens = batch.n_tokens; const int64_t n_tokens = batch.n_tokens;
GGML_ASSERT(lctx.inp_mean); GGML_ASSERT(lctx.inp_mean);
@ -13872,7 +13872,7 @@ static void llama_set_inputs(llama_context & lctx, const llama_batch & batch) {
} }
} }
if (cparams.pooling_type == LLAMA_POOLING_TYPE_CLS) { if (cparams.embeddings && cparams.pooling_type == LLAMA_POOLING_TYPE_CLS) {
const int64_t n_tokens = batch.n_tokens; const int64_t n_tokens = batch.n_tokens;
GGML_ASSERT(lctx.inp_cls); GGML_ASSERT(lctx.inp_cls);
@ -13893,7 +13893,7 @@ static void llama_set_inputs(llama_context & lctx, const llama_batch & batch) {
} }
} }
if (cparams.pooling_type == LLAMA_POOLING_TYPE_LAST) { if (cparams.embeddings && cparams.pooling_type == LLAMA_POOLING_TYPE_LAST) {
const int64_t n_tokens = batch.n_tokens; const int64_t n_tokens = batch.n_tokens;
GGML_ASSERT(lctx.inp_cls); GGML_ASSERT(lctx.inp_cls);
@ -14181,14 +14181,15 @@ static int llama_decode_internal(
std::vector<llama_seq_id *> seq_id_arr; std::vector<llama_seq_id *> seq_id_arr;
std::vector<std::vector<llama_seq_id>> seq_id; std::vector<std::vector<llama_seq_id>> seq_id;
// this indicates we are doing pooled embedding, so we ignore batch.logits and output all tokens
const bool embd_pooled = cparams.embeddings && cparams.pooling_type != LLAMA_POOLING_TYPE_NONE;
// count outputs // count outputs
if (cparams.embeddings && cparams.pooling_type != LLAMA_POOLING_TYPE_NONE) { if (batch_all.logits && !embd_pooled) {
n_outputs = n_tokens_all;
} else if (batch_all.logits) {
for (uint32_t i = 0; i < n_tokens_all; ++i) { for (uint32_t i = 0; i < n_tokens_all; ++i) {
n_outputs += batch_all.logits[i] != 0; n_outputs += batch_all.logits[i] != 0;
} }
} else if (lctx.logits_all) { } else if (lctx.logits_all || embd_pooled) {
n_outputs = n_tokens_all; n_outputs = n_tokens_all;
} else { } else {
// keep last output only // keep last output only
@ -14234,7 +14235,7 @@ static int llama_decode_internal(
{ {
int32_t n_outputs_new = 0; int32_t n_outputs_new = 0;
if (u_batch.logits) { if (u_batch.logits && !embd_pooled) {
for (uint32_t i = 0; i < n_tokens; i++) { for (uint32_t i = 0; i < n_tokens; i++) {
n_outputs_new += u_batch.logits[i] != 0; n_outputs_new += u_batch.logits[i] != 0;
} }
@ -18533,6 +18534,7 @@ struct llama_context_params llama_context_default_params() {
/*.n_threads_batch =*/ GGML_DEFAULT_N_THREADS, /*.n_threads_batch =*/ GGML_DEFAULT_N_THREADS,
/*.rope_scaling_type =*/ LLAMA_ROPE_SCALING_TYPE_UNSPECIFIED, /*.rope_scaling_type =*/ LLAMA_ROPE_SCALING_TYPE_UNSPECIFIED,
/*.pooling_type =*/ LLAMA_POOLING_TYPE_UNSPECIFIED, /*.pooling_type =*/ LLAMA_POOLING_TYPE_UNSPECIFIED,
/*.attention_type =*/ LLAMA_ATTENTION_TYPE_UNSPECIFIED,
/*.rope_freq_base =*/ 0.0f, /*.rope_freq_base =*/ 0.0f,
/*.rope_freq_scale =*/ 0.0f, /*.rope_freq_scale =*/ 0.0f,
/*.yarn_ext_factor =*/ -1.0f, /*.yarn_ext_factor =*/ -1.0f,
@ -18785,7 +18787,6 @@ struct llama_context * llama_new_context_with_model(
} }
cparams.yarn_attn_factor *= hparams.rope_attn_factor; cparams.yarn_attn_factor *= hparams.rope_attn_factor;
cparams.causal_attn = hparams.causal_attn;
if (cparams.pooling_type == LLAMA_POOLING_TYPE_UNSPECIFIED) { if (cparams.pooling_type == LLAMA_POOLING_TYPE_UNSPECIFIED) {
if (hparams.pooling_type == LLAMA_POOLING_TYPE_UNSPECIFIED) { if (hparams.pooling_type == LLAMA_POOLING_TYPE_UNSPECIFIED) {
@ -18795,6 +18796,12 @@ struct llama_context * llama_new_context_with_model(
} }
} }
if (params.attention_type == LLAMA_ATTENTION_TYPE_UNSPECIFIED) {
cparams.causal_attn = hparams.causal_attn;
} else {
cparams.causal_attn = params.attention_type == LLAMA_ATTENTION_TYPE_CAUSAL;
}
if (params.seed == LLAMA_DEFAULT_SEED) { if (params.seed == LLAMA_DEFAULT_SEED) {
params.seed = time(NULL); params.seed = time(NULL);
} }