llama: add support for small granite models

it works only for the small models 3b and 8b.

The convert-hf-to-gguf.py script uses the vocabulary size of the
granite models to detect granite and set the correct configuration.

Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>

convert-hf-to-gguf.py

@@ -1315,6 +1315,19 @@ def set_gguf_parameters(self):
                 self.gguf_writer.add_rope_scaling_type(gguf.RopeScalingType.LINEAR)
                 self.gguf_writer.add_rope_scaling_factor(self.hparams["rope_scaling"]["factor"])
 
+        # Apply to granite small models only
+        if self.hparams.get("vocab_size", 32000) == 49152:
+            self.gguf_writer.add_add_bos_token(False)
+            self.gguf_writer.add_rope_type(gguf.RopeType.NEOX)
+            self.gguf_writer.add_rope_scaling_type(gguf.RopeScalingType.NONE)
+
+        tokenizer_config_file = self.dir_model / 'tokenizer_config.json'
+        if tokenizer_config_file.is_file():
+            with open(tokenizer_config_file, "r", encoding="utf-8") as f:
+                tokenizer_config_json = json.load(f)
+                if "add_prefix_space" in tokenizer_config_json:
+                    self.gguf_writer.add_add_space_prefix(tokenizer_config_json["add_prefix_space"])
+
     @staticmethod
     def permute(weights: Tensor, n_head: int, n_head_kv: int | None):
         if n_head_kv is not None and n_head != n_head_kv:
@@ -1329,10 +1342,12 @@ def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iter
         n_head = self.hparams["num_attention_heads"]
         n_kv_head = self.hparams.get("num_key_value_heads")
 
-        if name.endswith("q_proj.weight"):
-            data_torch = LlamaModel.permute(data_torch, n_head, n_head)
-        if name.endswith("k_proj.weight"):
-            data_torch = LlamaModel.permute(data_torch, n_head, n_kv_head)
+        # Skip for granite models
+        if self.hparams.get("vocab_size", 32000) != 49152:
+            if name.endswith("q_proj.weight"):
+                data_torch = LlamaModel.permute(data_torch, n_head, n_head)
+            if name.endswith("k_proj.weight"):
+                data_torch = LlamaModel.permute(data_torch, n_head, n_kv_head)
 
         # process the experts separately
         if name.find("block_sparse_moe.experts") != -1:

llama.cpp

@@ -4001,7 +4001,9 @@ static void llm_load_hparams(
                     switch (hparams.n_layer) {
                         case 22: model.type = e_model::MODEL_1B; break;
                         case 26: model.type = e_model::MODEL_3B; break;
-                        case 32: model.type = hparams.n_vocab < 40000 ? e_model::MODEL_7B : e_model::MODEL_8B; break;
+                        // granite uses a vocab with len 49152
+                        case 32: model.type = hparams.n_vocab == 49152 ? e_model::MODEL_3B : (hparams.n_vocab < 40000 ? e_model::MODEL_7B : e_model::MODEL_8B); break;
+                        case 36: model.type = e_model::MODEL_8B; break; // granite
                         case 40: model.type = e_model::MODEL_13B; break;
                         case 48: model.type = e_model::MODEL_34B; break;
                         case 60: model.type = e_model::MODEL_30B; break;
@@ -4271,6 +4273,8 @@ static void llm_load_hparams(
                     case 30: model.type = e_model::MODEL_3B; break;
                     case 32: model.type = e_model::MODEL_7B; break;
                     case 40: model.type = e_model::MODEL_15B; break;
+                    case 52: model.type = e_model::MODEL_20B; break; // granite
+                    case 88: model.type = e_model::MODEL_34B; break; // granite
                     default: model.type = e_model::MODEL_UNKNOWN;
                 }
             } break;
@@ -4521,6 +4525,11 @@ static void llm_load_vocab(
         } else {
             if (tokenizer_model == "gpt2") {
                 vocab.type = LLAMA_VOCAB_TYPE_BPE;
+
+                const int add_space_prefix_keyidx = gguf_find_key(ctx, kv(LLM_KV_TOKENIZER_ADD_PREFIX).c_str());
+                if (add_space_prefix_keyidx != -1) {
+                    vocab.add_space_prefix = gguf_get_val_bool(ctx, add_space_prefix_keyidx);
+                }
             } else {
                 LLAMA_LOG_WARN("%s: unknown tokenizer: '%s'", __func__, tokenizer_model.c_str());
                 LLAMA_LOG_WARN("%s: using default tokenizer: 'llama'", __func__);