mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-11-07 09:57:00 +00:00
context : minor naming fix
This commit is contained in:
@@ -189,7 +189,7 @@ bool llama_context::apply_adapter_cvec(
|
||||
return cvec.apply(model, data, len, n_embd, il_start, il_end);
|
||||
}
|
||||
|
||||
llama_perf_context_data llama_context::get_perf() const {
|
||||
llama_perf_context_data llama_context::perf_get_data() const {
|
||||
llama_perf_context_data data = {};
|
||||
|
||||
data.t_start_ms = 1e-3 * t_start_us;
|
||||
|
||||
@@ -94,7 +94,6 @@ struct llama_context {
|
||||
//
|
||||
virtual int decode(llama_batch & inp_batch) = 0;
|
||||
|
||||
|
||||
// encode a batch of tokens by evaluating the encoder part of the transformer
|
||||
//
|
||||
// - lctx: llama context
|
||||
@@ -296,7 +295,7 @@ struct llama_context {
|
||||
|
||||
// perf
|
||||
|
||||
virtual llama_perf_context_data get_perf() const;
|
||||
virtual llama_perf_context_data perf_get_data() const;
|
||||
virtual void perf_reset();
|
||||
|
||||
// members
|
||||
@@ -326,8 +325,8 @@ protected:
|
||||
|
||||
bool has_evaluated_once = false;
|
||||
|
||||
mutable int64_t t_start_us;
|
||||
mutable int64_t t_load_us;
|
||||
mutable int64_t t_start_us = 0;
|
||||
mutable int64_t t_load_us = 0;
|
||||
mutable int64_t t_p_eval_us = 0;
|
||||
mutable int64_t t_eval_us = 0;
|
||||
|
||||
@@ -339,7 +338,8 @@ protected:
|
||||
};
|
||||
|
||||
// TODO: make implementation details private
|
||||
struct llama_context_unified : public llama_context {
|
||||
class llama_context_unified : public llama_context {
|
||||
public:
|
||||
struct batch_manager;
|
||||
|
||||
// TODO: tmp until llama-model starts implementing the graph build function
|
||||
|
||||
@@ -7854,7 +7854,7 @@ struct llama_perf_context_data llama_perf_context(const struct llama_context * c
|
||||
return data;
|
||||
}
|
||||
|
||||
data = ctx->get_perf();
|
||||
data = ctx->perf_get_data();
|
||||
|
||||
return data;
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user