cui-llama.rn 1.1.0 → 1.1.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/cpp/common.cpp +19 -6
- package/cpp/ggml-aarch64.c +6 -21
- package/cpp/ggml-metal.m +154 -26
- package/cpp/ggml.c +115 -195
- package/cpp/ggml.h +5 -7
- package/cpp/llama-impl.h +10 -4
- package/cpp/llama-sampling.cpp +16 -14
- package/cpp/llama.cpp +1048 -500
- package/cpp/llama.h +3 -0
- package/package.json +1 -1
package/cpp/llama.h
CHANGED
|
@@ -512,6 +512,9 @@ extern "C" {
|
|
|
512
512
|
// to the decoder to start generating output sequence. For other models, it returns -1.
|
|
513
513
|
LLAMA_API llama_token llama_model_decoder_start_token(const struct llama_model * model);
|
|
514
514
|
|
|
515
|
+
// Returns true if the model is recurrent (like Mamba, RWKV, etc.)
|
|
516
|
+
LLAMA_API bool llama_model_is_recurrent(const struct llama_model * model);
|
|
517
|
+
|
|
515
518
|
// Returns 0 on success
|
|
516
519
|
LLAMA_API uint32_t llama_model_quantize(
|
|
517
520
|
const char * fname_inp,
|