diff --git a/README.md b/README.md
index 973aac52..c984cffe 100644
--- a/README.md
+++ b/README.md
@@ -86,7 +86,7 @@ Choose your path:
 | Support Date | Model Name                                                 |
 | ------------ | ---------------------------------------------------------- |
 | Day 0        | Qwen2.5 / Qwen2-VL / QwQ / QvQ / InternLM3 / MiniCPM-o-2.6 |
-| Day 1        | Llama 3 / GLM-4 / PaliGemma2                               |
+| Day 1        | Llama 3 / GLM-4 / Mistral Small / PaliGemma2               |
 
 ## Benchmark
 
@@ -236,7 +236,9 @@ Compared to ChatGLM's [P-Tuning](https://github.com/THUDM/ChatGLM2-6B/tree/main/
 | [LLaVA-NeXT-Video](https://huggingface.co/llava-hf)               | 7B/34B                           | llava_next_video |
 | [MiniCPM](https://huggingface.co/openbmb)                         | 1B/2B/4B                         | cpm/cpm3         |
 | [MiniCPM-o-2.6/MiniCPM-V-2.6](https://huggingface.co/openbmb)     | 8B                               | minicpm_v        |
+| [Ministral/Mistral-Nemo](https://huggingface.co/mistralai)        | 8B/12B                           | ministral        |
 | [Mistral/Mixtral](https://huggingface.co/mistralai)               | 7B/8x7B/8x22B                    | mistral          |
+| [Mistral Small](https://huggingface.co/mistralai)                 | 24B                              | mistral_small    |
 | [OLMo](https://huggingface.co/allenai)                            | 1B/7B                            | -                |
 | [PaliGemma/PaliGemma2](https://huggingface.co/google)             | 3B/10B/28B                       | paligemma        |
 | [Phi-1.5/Phi-2](https://huggingface.co/microsoft)                 | 1.3B/2.7B                        | -                |
diff --git a/README_zh.md b/README_zh.md
index 079a272a..2157d11b 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -88,7 +88,7 @@ https://github.com/user-attachments/assets/e6ce34b0-52d5-4f3e-a830-592106c4c272
 | 适配时间      | 模型名称                                                    |
 | ------------ | ---------------------------------------------------------- |
 | Day 0        | Qwen2.5 / Qwen2-VL / QwQ / QvQ / InternLM3 / MiniCPM-o-2.6 |
-| Day 1        | Llama 3 / GLM-4 / PaliGemma2                               |
+| Day 1        | Llama 3 / GLM-4 / Mistral Small / PaliGemma2               |
 
 ## 性能指标
 
@@ -238,7 +238,9 @@ https://github.com/user-attachments/assets/e6ce34b0-52d5-4f3e-a830-592106c4c272
 | [LLaVA-NeXT-Video](https://huggingface.co/llava-hf)               | 7B/34B                           | llava_next_video |
 | [MiniCPM](https://huggingface.co/openbmb)                         | 1B/2B/4B                         | cpm/cpm3         |
 | [MiniCPM-o-2.6/MiniCPM-V-2.6](https://huggingface.co/openbmb)     | 8B                               | minicpm_v        |
+| [Ministral/Mistral-Nemo](https://huggingface.co/mistralai)        | 8B/12B                           | ministral        |
 | [Mistral/Mixtral](https://huggingface.co/mistralai)               | 7B/8x7B/8x22B                    | mistral          |
+| [Mistral Small](https://huggingface.co/mistralai)                 | 24B                              | mistral_small    |
 | [OLMo](https://huggingface.co/allenai)                            | 1B/7B                            | -                |
 | [PaliGemma/PaliGemma2](https://huggingface.co/google)             | 3B/10B/28B                       | paligemma        |
 | [Phi-1.5/Phi-2](https://huggingface.co/microsoft)                 | 1.3B/2.7B                        | -                |
diff --git a/requirements.txt b/requirements.txt
index 903aa1c3..d57ed94b 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,5 +1,5 @@
 transformers>=4.41.2,<=4.45.2;python_version<'3.10'
-transformers>=4.41.2,<=4.48.1,!=4.46.*,!=4.47.*,!=4.48.0;python_version>='3.10'
+transformers>=4.41.2,<=4.48.2,!=4.46.*,!=4.47.*,!=4.48.0;python_version>='3.10'
 datasets>=2.16.0,<=3.2.0
 accelerate>=0.34.0,<=1.2.1
 peft>=0.11.1,<=0.12.0
diff --git a/src/llamafactory/__init__.py b/src/llamafactory/__init__.py
index 0c3363c6..0b5d8cce 100644
--- a/src/llamafactory/__init__.py
+++ b/src/llamafactory/__init__.py
@@ -20,7 +20,7 @@ Level:
 
 Dependency graph:
   main:
-    transformers>=4.41.2,<=4.48.1,!=4.46.*,!=4.47.*,!=4.48.0
+    transformers>=4.41.2,<=4.48.2,!=4.46.*,!=4.47.*,!=4.48.0
     datasets>=2.16.0,<=3.2.0
     accelerate>=0.34.0,<=1.2.1
     peft>=0.11.1,<=0.12.0
@@ -30,7 +30,7 @@ Dependency graph:
   longlora:
     transformers>=4.41.2,<4.48.0
   packing:
-    transformers>=4.43.0,<=4.48.1
+    transformers>=4.43.0,<=4.48.2
 
 Disable version checking: DISABLE_VERSION_CHECK=1
 Enable VRAM recording: RECORD_VRAM=1
diff --git a/src/llamafactory/chat/hf_engine.py b/src/llamafactory/chat/hf_engine.py
index 57bbb405..7b5e3bbe 100644
--- a/src/llamafactory/chat/hf_engine.py
+++ b/src/llamafactory/chat/hf_engine.py
@@ -183,8 +183,8 @@ class HuggingfaceEngine(BaseEngine):
 
         if getattr(model.config, "model_type", None) in ["minicpmv", "minicpmo"]:
             gen_kwargs["input_ids"] = inputs
-            del gen_kwargs["image_sizes"]
             gen_kwargs["tokenizer"] = tokenizer
+            gen_kwargs.pop("image_sizes", None)
 
         return gen_kwargs, prompt_length
 
diff --git a/src/llamafactory/data/mm_plugin.py b/src/llamafactory/data/mm_plugin.py
index 00945923..eb8e7e5c 100644
--- a/src/llamafactory/data/mm_plugin.py
+++ b/src/llamafactory/data/mm_plugin.py
@@ -319,7 +319,7 @@ class LlavaNextPlugin(BasePlugin):
                 if self.expand_mm_tokens:
                     orig_height, orig_width = next(image_sizes)
                     image_seqlen = processor._get_number_of_features(orig_height, orig_width, height, width)
-                    if getattr(processor, "vision_feature_select_strategy") == "default":
+                    if getattr(processor, "vision_feature_select_strategy", "default") == "default":
                         image_seqlen -= 1
                 else:
                     image_seqlen = 1
@@ -370,7 +370,7 @@ class LlavaNextVideoPlugin(BasePlugin):
                     if self.expand_mm_tokens:
                         orig_height, orig_width = next(image_sizes)
                         image_seqlen = processor._get_number_of_features(orig_height, orig_width, height, width)
-                        if getattr(processor, "vision_feature_select_strategy") == "default":
+                        if getattr(processor, "vision_feature_select_strategy", "default") == "default":
                             image_seqlen -= 1
                     else:
                         image_seqlen = 1
@@ -915,7 +915,7 @@ class VideoLlavaPlugin(BasePlugin):
 
                 image_seqlen = (height // processor.patch_size) * (width // processor.patch_size) + 1
                 video_seqlen = image_seqlen * num_frames
-                if getattr(processor, "vision_feature_select_strategy") == "default":
+                if getattr(processor, "vision_feature_select_strategy", "default") == "default":
                     image_seqlen -= 1
             else:
                 image_seqlen, video_seqlen = 1, 1
diff --git a/src/llamafactory/data/template.py b/src/llamafactory/data/template.py
index 5b775db7..e87643da 100644
--- a/src/llamafactory/data/template.py
+++ b/src/llamafactory/data/template.py
@@ -220,6 +220,7 @@ def _register_template(
     replace_eos: bool = False,
     replace_jinja_template: bool = False,
     mm_plugin: "BasePlugin" = get_mm_plugin(name="base"),
+    fuse_system_into_user: bool = False,
 ) -> None:
     r"""
     Registers a chat template.
@@ -242,7 +243,7 @@ def _register_template(
     )
     ```
     """
-    template_class = Llama2Template if any(k in name for k in ("llama2", "mistral", "pixtral")) else Template
+    template_class = Llama2Template if fuse_system_into_user else Template
     default_slots = ["{{content}}"] if efficient_eos else ["{{content}}", {"eos_token"}]
     default_user_formatter = StringFormatter(slots=["{{content}}"])
     default_assistant_formatter = StringFormatter(slots=default_slots)
@@ -751,6 +752,7 @@ _register_template(
     name="llama2",
     format_user=StringFormatter(slots=[{"bos_token"}, "[INST] {{content}} [/INST]"]),
     format_system=StringFormatter(slots=["<<SYS>>\n{{content}}\n<</SYS>>\n\n"]),
+    fuse_system_into_user=True,
 )
 
 
@@ -760,6 +762,7 @@ _register_template(
     format_user=StringFormatter(slots=[{"bos_token"}, "[INST] {{content}} [/INST]"]),
     format_system=StringFormatter(slots=["<<SYS>>\n{{content}}\n<</SYS>>\n\n"]),
     default_system="You are a helpful assistant. 你是一个乐于助人的助手。",
+    fuse_system_into_user=True,
 )
 
 
@@ -878,11 +881,12 @@ _register_template(
     format_user=StringFormatter(slots=["[INST] {{content}}[/INST]"]),
     format_assistant=StringFormatter(slots=[" {{content}}", {"eos_token"}]),
     format_system=StringFormatter(slots=["{{content}}\n\n"]),
-    format_function=FunctionFormatter(slots=["[TOOL_CALLS] ", "{{content}}", {"eos_token"}], tool_format="mistral"),
+    format_function=FunctionFormatter(slots=["[TOOL_CALLS] {{content}}", {"eos_token"}], tool_format="mistral"),
     format_observation=StringFormatter(slots=["""[TOOL_RESULTS] {"content": {{content}}}[/TOOL_RESULTS]"""]),
     format_tools=ToolFormatter(tool_format="mistral"),
     format_prefix=EmptyFormatter(slots=[{"bos_token"}]),
     mm_plugin=get_mm_plugin(name="llava_next", image_token="<image>"),
+    fuse_system_into_user=True,
 )
 
 
@@ -932,11 +936,12 @@ _register_template(
     format_user=StringFormatter(slots=["[INST] {{content}}[/INST]"]),
     format_assistant=StringFormatter(slots=[" {{content}}", {"eos_token"}]),
     format_system=StringFormatter(slots=["{{content}}\n\n"]),
-    format_function=FunctionFormatter(slots=["[TOOL_CALLS] ", "{{content}}", {"eos_token"}], tool_format="mistral"),
+    format_function=FunctionFormatter(slots=["[TOOL_CALLS] {{content}}", {"eos_token"}], tool_format="mistral"),
     format_observation=StringFormatter(slots=["""[TOOL_RESULTS] {"content": {{content}}}[/TOOL_RESULTS]"""]),
     format_tools=ToolFormatter(tool_format="mistral"),
     format_prefix=EmptyFormatter(slots=[{"bos_token"}]),
     mm_plugin=get_mm_plugin(name="llava_next_video", image_token="<image>", video_token="<video>"),
+    fuse_system_into_user=True,
 )
 
 
@@ -978,15 +983,42 @@ _register_template(
 )
 
 
+# mistral tokenizer v3 tekken
+_register_template(
+    name="ministral",
+    format_user=StringFormatter(slots=["[INST]{{content}}[/INST]"]),
+    format_system=StringFormatter(slots=["{{content}}\n\n"]),
+    format_function=FunctionFormatter(slots=["[TOOL_CALLS]{{content}}", {"eos_token"}], tool_format="mistral"),
+    format_observation=StringFormatter(slots=["""[TOOL_RESULTS]{"content": {{content}}}[/TOOL_RESULTS]"""]),
+    format_tools=ToolFormatter(tool_format="mistral"),
+    format_prefix=EmptyFormatter(slots=[{"bos_token"}]),
+    fuse_system_into_user=True,
+)
+
+
+# mistral tokenizer v3
 _register_template(
     name="mistral",
     format_user=StringFormatter(slots=["[INST] {{content}}[/INST]"]),
     format_assistant=StringFormatter(slots=[" {{content}}", {"eos_token"}]),
     format_system=StringFormatter(slots=["{{content}}\n\n"]),
-    format_function=FunctionFormatter(slots=["[TOOL_CALLS] ", "{{content}}", {"eos_token"}], tool_format="mistral"),
+    format_function=FunctionFormatter(slots=["[TOOL_CALLS] {{content}}", {"eos_token"}], tool_format="mistral"),
     format_observation=StringFormatter(slots=["""[TOOL_RESULTS] {"content": {{content}}}[/TOOL_RESULTS]"""]),
     format_tools=ToolFormatter(tool_format="mistral"),
     format_prefix=EmptyFormatter(slots=[{"bos_token"}]),
+    fuse_system_into_user=True,
+)
+
+
+# mistral tokenizer v7 tekken (copied from ministral)
+_register_template(
+    name="mistral_small",
+    format_user=StringFormatter(slots=["[INST]{{content}}[/INST]"]),
+    format_system=StringFormatter(slots=["[SYSTEM_PROMPT]{{content}}[/SYSTEM_PROMPT]"]),
+    format_function=FunctionFormatter(slots=["[TOOL_CALLS]{{content}}", {"eos_token"}], tool_format="mistral"),
+    format_observation=StringFormatter(slots=["""[TOOL_RESULTS]{"content": {{content}}}[/TOOL_RESULTS]"""]),
+    format_tools=ToolFormatter(tool_format="mistral"),
+    format_prefix=EmptyFormatter(slots=[{"bos_token"}]),
 )
 
 
@@ -1081,12 +1113,17 @@ _register_template(
 )
 
 
+# copied from ministral template
 _register_template(
     name="pixtral",
     format_user=StringFormatter(slots=["[INST]{{content}}[/INST]"]),
     format_system=StringFormatter(slots=["{{content}}\n\n"]),
+    format_function=FunctionFormatter(slots=["[TOOL_CALLS]{{content}}", {"eos_token"}], tool_format="mistral"),
+    format_observation=StringFormatter(slots=["""[TOOL_RESULTS]{"content": {{content}}}[/TOOL_RESULTS]"""]),
+    format_tools=ToolFormatter(tool_format="mistral"),
     format_prefix=EmptyFormatter(slots=[{"bos_token"}]),
     mm_plugin=get_mm_plugin(name="pixtral", image_token="[IMG]"),
+    fuse_system_into_user=True,
 )
 
 
diff --git a/src/llamafactory/extras/constants.py b/src/llamafactory/extras/constants.py
index 20d84634..d570cc7a 100644
--- a/src/llamafactory/extras/constants.py
+++ b/src/llamafactory/extras/constants.py
@@ -1201,23 +1201,44 @@ register_model_group(
 
 register_model_group(
     models={
-        "MiniCPM-o-2_6-Chat": {
+        "MiniCPM-o-2_6": {
             DownloadSource.DEFAULT: "openbmb/MiniCPM-o-2_6",
             DownloadSource.MODELSCOPE: "OpenBMB/MiniCPM-o-2_6",
         },
     },
     template="minicpm_v",
+    vision=True,
 )
 
 
 register_model_group(
     models={
-        "MiniCPM-V-2_6-Chat": {
+        "MiniCPM-V-2_6": {
             DownloadSource.DEFAULT: "openbmb/MiniCPM-V-2_6",
             DownloadSource.MODELSCOPE: "OpenBMB/MiniCPM-V-2_6",
         },
     },
     template="minicpm_v",
+    vision=True,
+)
+
+
+register_model_group(
+    models={
+        "Ministral-8B-Instruct-2410": {
+            DownloadSource.DEFAULT: "mistralai/Ministral-8B-Instruct-2410",
+            DownloadSource.MODELSCOPE: "mistralai/Ministral-8B-Instruct-2410",
+        },
+        "Mistral-Nemo-Base-2407": {
+            DownloadSource.DEFAULT: "mistralai/Mistral-Nemo-Base-2407",
+            DownloadSource.MODELSCOPE: "LLM-Research/Mistral-Nemo-Base-2407",
+        },
+        "Mistral-Nemo-Instruct-2407": {
+            DownloadSource.DEFAULT: "mistralai/Mistral-Nemo-Instruct-2407",
+            DownloadSource.MODELSCOPE: "AI-ModelScope/Mistral-Nemo-Instruct-2407",
+        },
+    },
+    template="ministral",
 )
 
 
@@ -1227,48 +1248,60 @@ register_model_group(
             DownloadSource.DEFAULT: "mistralai/Mistral-7B-v0.1",
             DownloadSource.MODELSCOPE: "AI-ModelScope/Mistral-7B-v0.1",
         },
-        "Mistral-7B-Instruct-v0.1": {
-            DownloadSource.DEFAULT: "mistralai/Mistral-7B-Instruct-v0.1",
-            DownloadSource.MODELSCOPE: "AI-ModelScope/Mistral-7B-Instruct-v0.1",
-        },
         "Mistral-7B-v0.2": {
             DownloadSource.DEFAULT: "alpindale/Mistral-7B-v0.2-hf",
             DownloadSource.MODELSCOPE: "AI-ModelScope/Mistral-7B-v0.2-hf",
         },
+        "Mistral-7B-v0.3": {
+            DownloadSource.DEFAULT: "mistralai/Mistral-7B-v0.3",
+            DownloadSource.MODELSCOPE: "LLM-Research/mistral-7b-v0.3",
+        },
+        "Mistral-7B-Instruct-v0.1": {
+            DownloadSource.DEFAULT: "mistralai/Mistral-7B-Instruct-v0.1",
+            DownloadSource.MODELSCOPE: "AI-ModelScope/Mistral-7B-Instruct-v0.1",
+        },
         "Mistral-7B-Instruct-v0.2": {
             DownloadSource.DEFAULT: "mistralai/Mistral-7B-Instruct-v0.2",
             DownloadSource.MODELSCOPE: "AI-ModelScope/Mistral-7B-Instruct-v0.2",
         },
-        "Mistral-7B-v0.3": {
-            DownloadSource.DEFAULT: "mistralai/Mistral-7B-v0.3",
-        },
         "Mistral-7B-Instruct-v0.3": {
             DownloadSource.DEFAULT: "mistralai/Mistral-7B-Instruct-v0.3",
             DownloadSource.MODELSCOPE: "LLM-Research/Mistral-7B-Instruct-v0.3",
         },
-        "Mistral-Nemo-Instruct-2407": {
-            DownloadSource.DEFAULT: "mistralai/Mistral-Nemo-Instruct-2407",
-            DownloadSource.MODELSCOPE: "AI-ModelScope/Mistral-Nemo-Instruct-2407",
-        },
     },
     template="mistral",
 )
 
 
+register_model_group(
+    models={
+        "Mistral-Small-24B-Base-2501": {
+            DownloadSource.DEFAULT: "mistralai/Mistral-Small-24B-Base-2501",
+            DownloadSource.MODELSCOPE: "mistralai/Mistral-Small-24B-Base-2501",
+        },
+        "Mistral-Small-24B-Instruct-2501": {
+            DownloadSource.DEFAULT: "mistralai/Mistral-Small-24B-Instruct-2501",
+            DownloadSource.MODELSCOPE: "mistralai/Mistral-Small-24B-Instruct-2501",
+        },
+    },
+    template="mistral_small",
+)
+
+
 register_model_group(
     models={
         "Mixtral-8x7B-v0.1": {
             DownloadSource.DEFAULT: "mistralai/Mixtral-8x7B-v0.1",
             DownloadSource.MODELSCOPE: "AI-ModelScope/Mixtral-8x7B-v0.1",
         },
-        "Mixtral-8x7B-v0.1-Instruct": {
-            DownloadSource.DEFAULT: "mistralai/Mixtral-8x7B-Instruct-v0.1",
-            DownloadSource.MODELSCOPE: "AI-ModelScope/Mixtral-8x7B-Instruct-v0.1",
-        },
         "Mixtral-8x22B-v0.1": {
             DownloadSource.DEFAULT: "mistralai/Mixtral-8x22B-v0.1",
             DownloadSource.MODELSCOPE: "AI-ModelScope/Mixtral-8x22B-v0.1",
         },
+        "Mixtral-8x7B-v0.1-Instruct": {
+            DownloadSource.DEFAULT: "mistralai/Mixtral-8x7B-Instruct-v0.1",
+            DownloadSource.MODELSCOPE: "AI-ModelScope/Mixtral-8x7B-Instruct-v0.1",
+        },
         "Mixtral-8x22B-v0.1-Instruct": {
             DownloadSource.DEFAULT: "mistralai/Mixtral-8x22B-Instruct-v0.1",
             DownloadSource.MODELSCOPE: "AI-ModelScope/Mixtral-8x22B-Instruct-v0.1",
diff --git a/src/llamafactory/extras/misc.py b/src/llamafactory/extras/misc.py
index fdca25a6..7a198a85 100644
--- a/src/llamafactory/extras/misc.py
+++ b/src/llamafactory/extras/misc.py
@@ -94,7 +94,7 @@ def check_dependencies() -> None:
     r"""
     Checks the version of the required packages.
     """
-    check_version("transformers>=4.41.2,<=4.48.1,!=4.46.0,!=4.46.1,!=4.46.2,!=4.46.3,!=4.47.0,!=4.47.1,!=4.48.0")
+    check_version("transformers>=4.41.2,<=4.48.2,!=4.46.0,!=4.46.1,!=4.46.2,!=4.46.3,!=4.47.0,!=4.47.1,!=4.48.0")
     check_version("datasets>=2.16.0,<=3.2.0")
     check_version("accelerate>=0.34.0,<=1.2.1")
     check_version("peft>=0.11.1,<=0.12.0")
diff --git a/src/llamafactory/model/model_utils/packing.py b/src/llamafactory/model/model_utils/packing.py
index fd96813b..089c0274 100644
--- a/src/llamafactory/model/model_utils/packing.py
+++ b/src/llamafactory/model/model_utils/packing.py
@@ -118,6 +118,6 @@ def configure_packing(model_args: "ModelArguments", is_trainable: bool) -> None:
     if not is_trainable or not model_args.block_diag_attn:
         return
 
-    check_version("transformers>=4.43.0,<=4.48.1")
+    check_version("transformers>=4.43.0,<=4.48.2")
     transformers.modeling_flash_attention_utils._get_unpad_data = get_unpad_data
     logger.info_rank0("Using block diagonal attention for sequence packing without cross-attention.")