llm : add Falcon support (#2717)

* llama : refactor GGUF constants into static maps

* llama : check if model architecture is known

* llama : refactor llama_model_load_internal()

* gguf : add KV constant maps

* llm : read arch-specific KVs

* convert : add dummy scores + types

* falcon : load tensor data (CPU only)

* llama : fix loading progress bar

* llama : add arch member to llama_model

* falcon : CPU inference working

* falcon : support non-40B models

* falcon : minor

* llama : minor updates

ggml-ci

* convert-falcon-hf-to-gguf.py : fix special token mapping

* llama.cpp : llama default UNK token = id 0

* llama.cpp : fix bpe tokenizer

* llama.cpp : fix the fix of bpe tokenizer

* ggml : pass eps to ggml_norm

* metal : implement RoPE (mode = 2) + avoid ggml_repeat

* ggml : ggml_repeat always creates new tensor

* falcon : copy-paste self-attention from LLaMA

* metal : print extra compute pipeline info

* falcon : minor changes (still chasing the Metal problem)

* llama.cpp : fix linefeed token

* metal : fix GELU kernel numerical stability by using precise::tanh

* metal : temporary workaround for the concurrency optimization bug

* falcon : add CUDA offloading (#2739)

* llama : better model naming and size reporting

* llama : prep new tokenizer support

* llama : advanced BPE tokenizer based on ggllm.cpp imlpementation

* llama : remove oboslete comment

ggml-ci

* common : remove obsolete BPE API + disable test-tokenizer-1

* llama : revert BPE special-case in llama_byte_to_token()

* cuda : add TODOs for RoPE NeoX implementation

* llama : default special tokens based on vocab type

* perplexity : add log for start of tokenization

---------

Co-authored-by: klosax <131523366+klosax@users.noreply.github.com>
Co-authored-by: slaren <slarengh@gmail.com>

1 files changed, 8 insertions, 6 deletions

diff --git a/examples/main/main.cpp b/examples/main/main.cpp
index 0a22f3c2..1393f0b0 100644
--- a/examples/main/main.cpp
+++ b/examples/main/main.cpp
@@ -43,7 +43,7 @@ static bool is_interacting = false;
 void sigint_handler(int signo) {
     if (signo == SIGINT) {
         if (!is_interacting) {
-            is_interacting=true;
+            is_interacting = true;
         } else {
             console::cleanup();
             printf("\n");
@@ -189,10 +189,12 @@ int main(int argc, char ** argv) {
         }
     }
 
+    const bool is_spm = llama_vocab_type(ctx) == LLAMA_VOCAB_TYPE_SPM;
+
     // tokenize the prompt
     std::vector<llama_token> embd_inp;
     if (params.interactive_first || params.instruct || !params.prompt.empty() || session_tokens.empty()) {
-        embd_inp = ::llama_tokenize(ctx, params.prompt, true);
+        embd_inp = ::llama_tokenize(ctx, params.prompt, is_spm);
     } else {
         embd_inp = session_tokens;
     }
@@ -208,9 +210,9 @@ int main(int argc, char ** argv) {
     int original_prompt_len = 0;
     if (ctx_guidance) {
         params.cfg_negative_prompt.insert(0, 1, ' ');
-        guidance_inp = ::llama_tokenize(ctx_guidance, params.cfg_negative_prompt, true);
+        guidance_inp = ::llama_tokenize(ctx_guidance, params.cfg_negative_prompt, is_spm);
 
-        std::vector<llama_token> original_inp = ::llama_tokenize(ctx, params.prompt, true);
+        std::vector<llama_token> original_inp = ::llama_tokenize(ctx, params.prompt, is_spm);
         original_prompt_len = original_inp.size();
         guidance_offset = (int)guidance_inp.size() - original_prompt_len;
     }
@@ -257,8 +259,8 @@ int main(int argc, char ** argv) {
     }
 
     // prefix & suffix for instruct mode
-    const auto inp_pfx = ::llama_tokenize(ctx, "\n\n### Instruction:\n\n", true);
-    const auto inp_sfx = ::llama_tokenize(ctx, "\n\n### Response:\n\n", false);
+    const auto inp_pfx = ::llama_tokenize(ctx, "\n\n### Instruction:\n\n", is_spm);
+    const auto inp_sfx = ::llama_tokenize(ctx, "\n\n### Response:\n\n",    false);
 
     // in instruct mode, we inject a prefix and a suffix to each input by the user
     if (params.instruct) {