docs : Minor cleanups (#19252)

* Update old URLs to github.com/ggml-org/

* Bump copyrights

src/llama-chat.cpp

@@ -233,7 +233,7 @@ int32_t llm_chat_apply_template(
     llm_chat_template tmpl,
     const std::vector<const llama_chat_message *> & chat,
     std::string & dest, bool add_ass) {
-    // Taken from the research: https://github.com/ggerganov/llama.cpp/issues/5527
+    // Taken from the research: https://github.com/ggml-org/llama.cpp/issues/5527
     std::stringstream ss;
     if (tmpl == LLM_CHAT_TEMPLATE_CHATML) {
         // chatml template

src/llama-hparams.h

@@ -195,7 +195,7 @@ struct llama_hparams {
     uint32_t n_deepstack_layers = 0;
 
     // needed by encoder-decoder models (e.g. T5, FLAN-T5)
-    // ref: https://github.com/ggerganov/llama.cpp/pull/8141
+    // ref: https://github.com/ggml-org/llama.cpp/pull/8141
     llama_token dec_start_token_id = LLAMA_TOKEN_NULL;
     uint32_t    dec_n_layer        = 0;
 

src/llama-vocab.cpp

@@ -90,7 +90,7 @@ static_assert(std::is_trivially_copyable<llm_symbol>::value, "llm_symbol is not
 //
 // SPM tokenizer
 // original implementation:
-// https://github.com/ggerganov/llama.cpp/commit/074bea2eb1f1349a0118239c4152914aecaa1be4
+// https://github.com/ggml-org/llama.cpp/commit/074bea2eb1f1349a0118239c4152914aecaa1be4
 //
 
 struct llm_bigram_spm {
@@ -285,7 +285,7 @@ struct llm_tokenizer_bpe : llm_tokenizer {
                     // original regex from tokenizer.json
                     //"(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?\\p{L}+|\\p{N}{1,3}| ?[^\\s\\p{L}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
 
-                    // adapted: https://github.com/ggerganov/llama.cpp/pull/6920#issuecomment-2080233989
+                    // adapted: https://github.com/ggml-org/llama.cpp/pull/6920#issuecomment-2080233989
                     "(?:'[sS]|'[tT]|'[rR][eE]|'[vV][eE]|'[mM]|'[lL][lL]|'[dD])|[^\\r\\n\\p{L}\\p{N}]?\\p{L}+|\\p{N}{1,3}| ?[^\\s\\p{L}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
                 };
                 break;
@@ -2390,7 +2390,7 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {
 
         // maintain a list of tokens that cause end-of-generation
         // this is currently determined based on the token text, which is obviously not ideal
-        // ref: https://github.com/ggerganov/llama.cpp/issues/9606
+        // ref: https://github.com/ggml-org/llama.cpp/issues/9606
         special_eog_ids.clear();
 
         if (special_fim_pad_id != LLAMA_TOKEN_NULL && special_eog_ids.count(special_fim_pad_id) == 0) {
@@ -3079,7 +3079,7 @@ std::vector<llama_token> llama_vocab::impl::tokenize(
 }
 
 int32_t llama_vocab::impl::token_to_piece(llama_token token, char * buf, int32_t length, int32_t lstrip, bool special) const {
-    // ref: https://github.com/ggerganov/llama.cpp/pull/7587#discussion_r1620983843
+    // ref: https://github.com/ggml-org/llama.cpp/pull/7587#discussion_r1620983843
     static const int attr_special = LLAMA_TOKEN_ATTR_UNKNOWN | LLAMA_TOKEN_ATTR_CONTROL;
     const llama_token_attr attr = token_get_attr(token);
     if (!special && (attr & attr_special)) {

src/models/deepseek2.cpp

@@ -14,7 +14,7 @@ llm_build_deepseek2::llm_build_deepseek2(const llama_model & model, const llm_gr
     const uint32_t kv_lora_rank = hparams.n_lora_kv;
 
     // We have to pre-scale kq_scale and attn_factor to make the YaRN RoPE work correctly.
-    // See https://github.com/ggerganov/llama.cpp/discussions/7416 for detailed explanation.
+    // See https://github.com/ggml-org/llama.cpp/discussions/7416 for detailed explanation.
     // And also: https://github.com/ggml-org/llama.cpp/pull/17945 [TAG_DEEPSEEK2_YARN_LOG_MUL_FIX]
 
     // first cancel the adjustment from llama_hparams::yarn_attn_factor_adjust to get the original attn_factor