diff options
| author | slaren <slarengh@gmail.com> | 2023-09-28 21:42:38 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-09-28 22:42:38 +0300 |
| commit | 16bc66d9479edd5ee12ec734973554d4493c5dfa (patch) | |
| tree | 4cca787ebd86dd55fd176d27112117c74e9b34c6 /common/train.cpp | |
| parent | 0512d66670de3f650c579519833c085014b0f200 (diff) | |
llama.cpp : split llama_context_params into model and context params (#3301)
* llama.cpp : split llama_context_params into model and context params
ggml-ci
* fix metal build
* fix freq_base/scale default to model value
* llama-bench : keep the same model between tests when possible
* move n_threads to llama_context_params, add n_threads_batch
* fix mpi build
* remove kv_size(), cuda scratch fixes
* remove low-vram option
* add n_threads_batch to system info, refactor to get_system_info()
* add documentation about --threads-batch to the READMEs
* llama-bench fix
* main : fix rope freq/scale warning
* llama.cpp : add llama_get_model
common : add llama_tokenize from model
* remove duplicated ctx/model functions
ggml-ci
* cuda : print total VRAM used
Diffstat (limited to 'common/train.cpp')
| -rw-r--r-- | common/train.cpp | 10 |
1 files changed, 5 insertions, 5 deletions
diff --git a/common/train.cpp b/common/train.cpp index 4a128096..35a4cf9e 100644 --- a/common/train.cpp +++ b/common/train.cpp @@ -858,7 +858,7 @@ size_t tokenize_file( out_tokens.resize(buf.size() + n_max_tokens_overhead); int n_tokens = llama_tokenize( - lctx, + llama_get_model(lctx), buf.data(), (int) buf.size(), out_tokens.data(), @@ -867,7 +867,7 @@ size_t tokenize_file( if (n_tokens < 0) { out_tokens.resize(-n_tokens); n_tokens = llama_tokenize( - lctx, + llama_get_model(lctx), buf.data(), (int) buf.size(), out_tokens.data(), @@ -920,7 +920,7 @@ size_t tokenize_file( size_t found_max_sample_size = 0; size_t max_token_text_size = 0; - int n_vocab = llama_n_vocab(lctx); + int n_vocab = llama_n_vocab(llama_get_model(lctx)); for (llama_token token=0; token < n_vocab; ++token) { max_token_text_size = std::max( max_token_text_size, @@ -961,7 +961,7 @@ size_t tokenize_file( // tokenize the sample tok_sample.resize(buf_sample.size() + n_max_tokens_overhead); - int n_tokens = llama_tokenize(lctx, + int n_tokens = llama_tokenize(llama_get_model(lctx), buf_sample.data(), (int) buf_sample.size(), tok_sample.data(), @@ -969,7 +969,7 @@ size_t tokenize_file( false); if (n_tokens < 0) { tok_sample.resize(-n_tokens); - n_tokens = llama_tokenize(lctx, + n_tokens = llama_tokenize(llama_get_model(lctx), buf_sample.data(), (int) buf_sample.size(), tok_sample.data(), |