Skip to content

GitLab

Unverified Commit 2f80c556 authored by q.yao's avatar q.yao Committed by GitHub
Browse files

fix typo (#769)

parent 7868cea5
Show whitespace changes
Inline Side-by-side
Showing with 3 additions and 3 deletions
src/turbomind/models/llama/LlamaV2.cc
View file @ 2f80c556
......@@ -218,7 +218,7 @@ void LlamaV2<T>::embeddingLookup(T* embeddings, const int* token_ids_buf, int ba
}
template<typename T>
void LlamaV2<T>::contextDecode(T* deocder_output,
void LlamaV2<T>::contextDecode(T* decoder_output,
uintptr_t* k_cache_ptr,
uintptr_t* v_cache_ptr,
void** tmp_k_ptrs,
......@@ -280,7 +280,7 @@ void LlamaV2<T>::contextDecode(T* deocder_output,
{"value_cache", {MEMORY_GPU, TYPE_UINT64, {bsz}, v_cache_ptr}},
{"tmp_k", {MEMORY_GPU, TYPE_UINT64, {bsz}, tmp_k_ptrs}},
{"tmp_v", {MEMORY_GPU, TYPE_UINT64, {bsz}, tmp_v_ptrs}},
{"last_token_hidden_units", {MEMORY_GPU, dtype, {bsz, hidden_units_}, deocder_output}}};
{"last_token_hidden_units", {MEMORY_GPU, dtype, {bsz, hidden_units_}, decoder_output}}};
context_decoder_->forward(&decoder_output_tensors, &decoder_input_tensors, &weights_->decoder_layer_weights);
......
src/turbomind/models/llama/LlamaV2.h
View file @ 2f80c556
......@@ -113,7 +113,7 @@ private:
void embeddingLookup(T* embeddings, const int* token_ids_buf, int batch_size, int step);
void contextDecode(T* deocder_output,
void contextDecode(T* decoder_output,
uintptr_t* k_block_ptrs,
uintptr_t* v_block_ptrs,
void** k_tmp_ptrs,
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment