fix old 9b err & remove some note debug log

Author: piDack

convert_hf_to_gguf.py

@@ -281,10 +281,8 @@ def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iter
         return [(self.map_tensor_name(name), data_torch)]
 
     def tensor_force_quant(self, name: str, new_name: str, bid: int | None, n_dims: int) -> gguf.GGMLQuantizationType | bool:
-        # 删除未使用的参数，避免警告或错误
         del name, new_name, bid, n_dims  # unused
 
-        # 返回False，表示不进行量化
         return False
 
     # some models need extra generated tensors (like rope_freqs)
@@ -5051,7 +5049,6 @@ def set_vocab(self):
         special_vocab._set_special_token("unk", tokenizer.get_added_vocab()["<|endoftext|>"])
         # exclude glm-edge 1.5B & 4B
         if self.hparams.get("partial_rotary_factor", 1.0) == 0.5:
-            print("add bos in model")
             special_vocab._set_special_token("bos", tokenizer.get_added_vocab()["[gMASK]"])
         special_vocab.add_to_gguf(self.gguf_writer)
 

src/llama-model.cpp

@@ -3467,7 +3467,6 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
                     // output
                     output_norm   = create_tensor(tn(LLM_TENSOR_OUTPUT_NORM, "weight"), {n_embd}, 0);
                     output        = create_tensor(tn(LLM_TENSOR_OUTPUT,      "weight"), {n_embd, n_vocab}, 0);
-                    // if output is NULL, init from the input tok embed
                     for (int i = 0; i < n_layer; ++i) {
                         auto & layer = layers[i];
 
@@ -3494,7 +3493,7 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
 
                         layer.ffn_down   = create_tensor(tn(LLM_TENSOR_FFN_DOWN, "weight", i), {n_ff, n_embd}, 0);
 
-                        layer.ffn_post_norm  = create_tensor(tn(LLM_TENSOR_FFN_POST_NORM, "weight", i), {n_embd}, 0);
+                        layer.ffn_post_norm  = create_tensor(tn(LLM_TENSOR_FFN_POST_NORM, "weight", i), {n_embd}, TENSOR_NOT_REQUIRED);
                     }
                 } break;
             case LLM_ARCH_NEMOTRON: