update 0718

src/llamafactory/extras/misc.py

@@ -23,6 +23,7 @@ from typing import TYPE_CHECKING, Any, Literal, Union
 import torch
 import torch.distributed as dist
 import transformers.dynamic_module_utils
+from huggingface_hub.utils import WeakFileLock
 from transformers import InfNanRemoveLogitsProcessor, LogitsProcessorList
 from transformers.dynamic_module_utils import get_relative_imports
 from transformers.utils import (
@@ -35,7 +36,6 @@ from transformers.utils import (
 from transformers.utils.versions import require_version
 
 from . import logging
-from .packages import is_transformers_version_greater_than
 
 
 _is_fp16_available = is_torch_npu_available() or is_torch_cuda_available()
@@ -94,15 +94,11 @@ def check_version(requirement: str, mandatory: bool = False) -> None:
 
 def check_dependencies() -> None:
     r"""Check the version of the required packages."""
-    check_version(
-        "transformers>=4.45.0,<=4.52.4,!=4.46.0,!=4.46.1,!=4.46.2,!=4.46.3,!=4.47.0,!=4.47.1,!=4.48.0,!=4.52.0"
-    )
+    check_version("transformers>=4.49.0,<=4.52.4,!=4.52.0")
     check_version("datasets>=2.16.0,<=3.6.0")
-    check_version("accelerate>=0.34.0,<=1.7.0")
+    check_version("accelerate>=1.3.0,<=1.7.0")
     check_version("peft>=0.14.0,<=0.15.2")
     check_version("trl>=0.8.6,<=0.9.6")
-    if is_transformers_version_greater_than("4.46.0") and not is_transformers_version_greater_than("4.48.1"):
-        logger.warning_rank0_once("There are known bugs in transformers v4.46.0-v4.48.0, please use other versions.")
 
 
 def calculate_tps(dataset: list[dict[str, Any]], metrics: dict[str, float], stage: Literal["sft", "rm"]) -> float:
@@ -182,8 +178,22 @@ def get_logits_processor() -> "LogitsProcessorList":
     return logits_processor
 
 
+def get_current_memory() -> tuple[int, int]:
+    r"""Get the available and total memory for the current device (in Bytes)."""
+    if is_torch_xpu_available():
+        return torch.xpu.mem_get_info()
+    elif is_torch_npu_available():
+        return torch.npu.mem_get_info()
+    elif is_torch_mps_available():
+        return torch.mps.current_allocated_memory(), torch.mps.recommended_max_memory()
+    elif is_torch_cuda_available():
+        return torch.cuda.mem_get_info()
+    else:
+        return 0, -1
+
+
 def get_peak_memory() -> tuple[int, int]:
-    r"""Get the peak memory usage for the current device (in Bytes)."""
+    r"""Get the peak memory usage (allocated, reserved) for the current device (in Bytes)."""
     if is_torch_xpu_available():
         return torch.xpu.max_memory_allocated(), torch.xpu.max_memory_reserved()
     elif is_torch_npu_available():
@@ -193,7 +203,7 @@ def get_peak_memory() -> tuple[int, int]:
     elif is_torch_cuda_available():
         return torch.cuda.max_memory_allocated(), torch.cuda.max_memory_reserved()
     else:
-        return 0, 0
+        return 0, -1
 
 
 def has_tokenized_data(path: "os.PathLike") -> bool:
@@ -259,26 +269,37 @@ def try_download_model_from_other_hub(model_args: "ModelArguments") -> str:
         return model_args.model_name_or_path
 
     if use_modelscope():
-        check_version("modelscope>=1.11.0", mandatory=True)
+        check_version("modelscope>=1.14.0", mandatory=True)
         from modelscope import snapshot_download  # type: ignore
+        from modelscope.hub.api import HubApi  # type: ignore
+
+        if model_args.ms_hub_token:
+            api = HubApi()
+            api.login(model_args.ms_hub_token)
 
         revision = "master" if model_args.model_revision == "main" else model_args.model_revision
-        return snapshot_download(
+        with WeakFileLock(os.path.abspath(os.path.expanduser("~/.cache/llamafactory/modelscope.lock"))):
+            model_path = snapshot_download(
                 model_args.model_name_or_path,
                 revision=revision,
                 cache_dir=model_args.cache_dir,
             )
 
+        return model_path
+
     if use_openmind():
         check_version("openmind>=0.8.0", mandatory=True)
         from openmind.utils.hub import snapshot_download  # type: ignore
 
-        return snapshot_download(
+        with WeakFileLock(os.path.abspath(os.path.expanduser("~/.cache/llamafactory/openmind.lock"))):
+            model_path = snapshot_download(
                 model_args.model_name_or_path,
                 revision=model_args.model_revision,
                 cache_dir=model_args.cache_dir,
             )
 
+        return model_path
+
 
 def use_modelscope() -> bool:
     return is_env_enabled("USE_MODELSCOPE_HUB")
@@ -305,5 +326,5 @@ def fix_proxy(ipv6_enabled: bool = False) -> None:
     r"""Fix proxy settings for gradio ui."""
     os.environ["no_proxy"] = "localhost,127.0.0.1,0.0.0.0"
     if ipv6_enabled:
-        for name in ("http_proxy", "https_proxy", "HTTP_PROXY", "HTTPS_PROXY"):
-            os.environ.pop(name, None)
+        os.environ.pop("http_proxy", None)
+        os.environ.pop("HTTP_PROXY", None)

src/llamafactory/hparams/parser.py

@@ -15,7 +15,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import json
 import os
 import sys
 from pathlib import Path
@@ -23,7 +22,6 @@ from typing import Any, Optional, Union
 
 import torch
 import transformers
-import yaml
 from omegaconf import OmegaConf
 from transformers import HfArgumentParser
 from transformers.integrations import is_deepspeed_zero3_enabled
@@ -62,11 +60,11 @@ def read_args(args: Optional[Union[dict[str, Any], list[str]]] = None) -> Union[
 
     if sys.argv[1].endswith(".yaml") or sys.argv[1].endswith(".yml"):
         override_config = OmegaConf.from_cli(sys.argv[2:])
-        dict_config = yaml.safe_load(Path(sys.argv[1]).absolute().read_text())
+        dict_config = OmegaConf.load(Path(sys.argv[1]).absolute())
         return OmegaConf.to_container(OmegaConf.merge(dict_config, override_config))
     elif sys.argv[1].endswith(".json"):
         override_config = OmegaConf.from_cli(sys.argv[2:])
-        dict_config = json.loads(Path(sys.argv[1]).absolute().read_text())
+        dict_config = OmegaConf.load(Path(sys.argv[1]).absolute())
         return OmegaConf.to_container(OmegaConf.merge(dict_config, override_config))
     else:
         return sys.argv[1:]
@@ -166,6 +164,9 @@ def _check_extra_dependencies(
     if finetuning_args.use_adam_mini:
         check_version("adam-mini", mandatory=True)
 
+    if finetuning_args.use_swanlab:
+        check_version("swanlab", mandatory=True)
+
     if finetuning_args.plot_loss:
         check_version("matplotlib", mandatory=True)
 
@@ -348,6 +349,9 @@ def get_train_args(args: Optional[Union[dict[str, Any], list[str]]] = None) -> _
         # https://github.com/huggingface/transformers/blob/v4.50.0/src/transformers/trainer.py#L782
         training_args.label_names = training_args.label_names or ["labels"]
 
+    if "swanlab" in training_args.report_to and finetuning_args.use_swanlab:
+        training_args.report_to.remove("swanlab")
+
     if (
         training_args.parallel_mode == ParallelMode.DISTRIBUTED
         and training_args.ddp_find_unused_parameters is None

src/llamafactory/model/adapter.py

@@ -188,7 +188,7 @@ def _setup_lora_tuning(
 
         if adapter_to_resume is not None:  # resume lora training
             if model_args.use_unsloth:
-                model = load_unsloth_peft_model(config, model_args, is_trainable=is_trainable)
+                model = load_unsloth_peft_model(config, model_args, finetuning_args, is_trainable=is_trainable)
             else:
                 model = PeftModel.from_pretrained(model, adapter_to_resume, is_trainable=is_trainable, **init_kwargs)
 

src/llamafactory/model/loader.py

@@ -19,6 +19,7 @@ import torch
 from transformers import (
     AutoConfig,
     AutoModelForCausalLM,
+    AutoModelForImageTextToText,
     AutoModelForSeq2SeqLM,
     AutoModelForTextToWaveform,
     AutoModelForVision2Seq,
@@ -29,7 +30,6 @@ from trl import AutoModelForCausalLMWithValueHead
 
 from ..extras import logging
 from ..extras.misc import count_parameters, skip_check_imports, try_download_model_from_other_hub
-from ..extras.packages import is_transformers_version_greater_than
 from .adapter import init_adapter
 from .model_utils.liger_kernel import apply_liger_kernel
 from .model_utils.misc import register_autoclass
@@ -39,10 +39,6 @@ from .model_utils.valuehead import load_valuehead_params
 from .patcher import patch_config, patch_model, patch_processor, patch_tokenizer, patch_valuehead_model
 
 
-if is_transformers_version_greater_than("4.46.0"):
-    from transformers import AutoModelForImageTextToText
-
-
 if TYPE_CHECKING:
     from transformers import PretrainedConfig, PreTrainedModel, PreTrainedTokenizer, ProcessorMixin
 
@@ -111,9 +107,8 @@ def load_tokenizer(model_args: "ModelArguments") -> "TokenizerModule":
             **init_kwargs,
         )
     except Exception as e:
-        raise OSError("Failed to load processor.") from e
-
-    patch_processor(processor, tokenizer, model_args)
+        logger.info_rank0(f"Failed to load processor: {e}.")
+        processor = None
 
     # Avoid load tokenizer, see:
     # https://github.com/huggingface/transformers/blob/v4.40.0/src/transformers/models/auto/processing_auto.py#L324
@@ -121,6 +116,9 @@ def load_tokenizer(model_args: "ModelArguments") -> "TokenizerModule":
         logger.debug("The loaded processor is not an instance of Processor. Dropping it.")
         processor = None
 
+    if processor is not None:
+        patch_processor(processor, tokenizer, model_args)
+
     return {"tokenizer": tokenizer, "processor": processor}
 
 
@@ -160,10 +158,7 @@ def load_model(
         else:
             if type(config) in AutoModelForVision2Seq._model_mapping.keys():  # image-text
                 load_class = AutoModelForVision2Seq
-            elif (
-                is_transformers_version_greater_than("4.46.0")
-                and type(config) in AutoModelForImageTextToText._model_mapping.keys()
-            ):  # image-text
+            elif type(config) in AutoModelForImageTextToText._model_mapping.keys():  # image-text
                 load_class = AutoModelForImageTextToText
             elif type(config) in AutoModelForSeq2SeqLM._model_mapping.keys():  # audio-text
                 load_class = AutoModelForSeq2SeqLM

src/llamafactory/model/model_utils/unsloth.py

@@ -80,12 +80,15 @@ def get_unsloth_peft_model(
 
 
 def load_unsloth_peft_model(
-    config: "PretrainedConfig", model_args: "ModelArguments", is_trainable: bool
+    config: "PretrainedConfig",
+    model_args: "ModelArguments",
+    finetuning_args: "FinetuningArguments",
+    is_trainable: bool,
 ) -> "PreTrainedModel":
     r"""Load peft model with unsloth. Used in both training and inference."""
     from unsloth import FastLanguageModel  # type: ignore
 
-    unsloth_kwargs = _get_unsloth_kwargs(config, model_args.adapter_name_or_path[0], model_args)
+    unsloth_kwargs = _get_unsloth_kwargs(config, model_args.adapter_name_or_path[0], model_args, finetuning_args)
     try:
         if not is_trainable:
             unsloth_kwargs["use_gradient_checkpointing"] = False

src/llamafactory/model/model_utils/valuehead.py

@@ -49,7 +49,7 @@ def load_valuehead_params(path_or_repo_id: str, model_args: "ModelArguments") ->
 
     try:
         vhead_file = cached_file(filename=V_HEAD_WEIGHTS_NAME, **kwargs)
-        return torch.load(vhead_file, map_location="cpu")
+        return torch.load(vhead_file, map_location="cpu", weights_only=True)
     except Exception as err:
         err_text = str(err)
 

src/llamafactory/model/model_utils/visual.py

@@ -204,6 +204,23 @@ _register_composite_model(
 )
 
 
+_register_composite_model(
+    model_type="gemma3n",
+    vision_model_keys=["vision_tower", "audio_tower"],
+    lora_conflict_keys=["timm_model", "subsample_conv_projection"],
+)
+
+
+# copied from qwen2vl
+_register_composite_model(
+    model_type="glm4v",
+    projector_key="visual.merger",
+    vision_model_keys=["visual.patch_embed", "visual.blocks"],
+    language_model_keys=["language_model", "lm_head"],
+    lora_conflict_keys=["patch_embed"],
+)
+
+
 _register_composite_model(
     model_type="internvl",
 )

src/llamafactory/model/patcher.py

@@ -178,6 +178,9 @@ def patch_model(
         resize_embedding_layer(model, tokenizer)
 
     if is_trainable:
+        if getattr(model.config, "model_type", None) == "gemma3n":
+            setattr(model_args, "disable_gradient_checkpointing", True)
+
         prepare_model_for_training(model, model_args)
         autocast_projector_dtype(model, model_args)
         add_z3_leaf_module(model)

src/llamafactory/train/callbacks.py

@@ -76,7 +76,7 @@ def fix_valuehead_checkpoint(
             state_dict: dict[str, torch.Tensor] = {key: f.get_tensor(key) for key in f.keys()}
     else:
         path_to_checkpoint = os.path.join(output_dir, WEIGHTS_NAME)
-        state_dict: dict[str, torch.Tensor] = torch.load(path_to_checkpoint, map_location="cpu")
+        state_dict: dict[str, torch.Tensor] = torch.load(path_to_checkpoint, map_location="cpu", weights_only=True)
 
     os.remove(path_to_checkpoint)
     decoder_state_dict, v_head_state_dict = {}, {}

src/llamafactory/webui/common.py

@@ -77,14 +77,19 @@ def load_config() -> dict[str, Union[str, dict[str, Any]]]:
         with open(_get_config_path(), encoding="utf-8") as f:
             return safe_load(f)
     except Exception:
-        return {"lang": None, "last_model": None, "path_dict": {}, "cache_dir": None}
+        return {"lang": None, "hub_name": None, "last_model": None, "path_dict": {}, "cache_dir": None}
 
 
-def save_config(lang: str, model_name: Optional[str] = None, model_path: Optional[str] = None) -> None:
+def save_config(
+    lang: str, hub_name: Optional[str] = None, model_name: Optional[str] = None, model_path: Optional[str] = None
+) -> None:
     r"""Save user config."""
     os.makedirs(DEFAULT_CACHE_DIR, exist_ok=True)
     user_config = load_config()
     user_config["lang"] = lang or user_config["lang"]
+    if hub_name:
+        user_config["hub_name"] = hub_name
+
     if model_name:
         user_config["last_model"] = model_name
 
@@ -247,7 +252,7 @@ def create_ds_config() -> None:
         "stage": 2,
         "allgather_partitions": True,
         "allgather_bucket_size": 5e8,
-        "overlap_comm": True,
+        "overlap_comm": False,
         "reduce_scatter": True,
         "reduce_bucket_size": 5e8,
         "contiguous_gradients": True,
@@ -262,7 +267,7 @@ def create_ds_config() -> None:
 
     ds_config["zero_optimization"] = {
         "stage": 3,
-        "overlap_comm": True,
+        "overlap_comm": False,
         "contiguous_gradients": True,
         "sub_group_size": 1e9,
         "reduce_bucket_size": "auto",

src/llamafactory/webui/components/__init__.py

@@ -15,6 +15,7 @@
 from .chatbot import create_chat_box
 from .eval import create_eval_tab
 from .export import create_export_tab
+from .footer import create_footer
 from .infer import create_infer_tab
 from .top import create_top
 from .train import create_train_tab
@@ -24,6 +25,7 @@ __all__ = [
     "create_chat_box",
     "create_eval_tab",
     "create_export_tab",
+    "create_footer",
     "create_infer_tab",
     "create_top",
     "create_train_tab",

src/llamafactory/webui/components/chatbot.py

@@ -12,6 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import inspect
 import json
 from typing import TYPE_CHECKING
 
@@ -50,7 +51,14 @@ def create_chat_box(
 ) -> tuple["Component", "Component", dict[str, "Component"]]:
     lang = engine.manager.get_elem_by_id("top.lang")
     with gr.Column(visible=visible) as chat_box:
-        chatbot = gr.Chatbot(type="messages", show_copy_button=True)
+        kwargs = {}
+        if "show_copy_button" in inspect.signature(gr.Chatbot.__init__).parameters:
+            kwargs["show_copy_button"] = True
+
+        if "resizable" in inspect.signature(gr.Chatbot.__init__).parameters:
+            kwargs["resizable"] = True
+
+        chatbot = gr.Chatbot(type="messages", **kwargs)
         messages = gr.State([])
         with gr.Row():
             with gr.Column(scale=4):

src/llamafactory/webui/components/footer.py

+# Copyright 2025 the LlamaFactory team.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import TYPE_CHECKING
+
+from ...extras.misc import get_current_memory
+from ...extras.packages import is_gradio_available
+
+
+if is_gradio_available():
+    import gradio as gr
+
+
+if TYPE_CHECKING:
+    from gradio.components import Component
+
+
+def get_device_memory() -> "gr.Slider":
+    free, total = get_current_memory()
+    if total != -1:
+        used = round((total - free) / (1024**3), 2)
+        total = round(total / (1024**3), 2)
+        return gr.Slider(minimum=0, maximum=total, value=used, step=0.01, visible=True)
+    else:
+        return gr.Slider(visible=False)
+
+
+def create_footer() -> dict[str, "Component"]:
+    with gr.Row():
+        device_memory = gr.Slider(visible=False, interactive=False)
+        timer = gr.Timer(value=5)
+
+    timer.tick(get_device_memory, outputs=[device_memory], queue=False)
+    return dict(device_memory=device_memory)

src/llamafactory/webui/components/top.py

@@ -16,9 +16,10 @@ from typing import TYPE_CHECKING
 
 from ...data import TEMPLATES
 from ...extras.constants import METHODS, SUPPORTED_MODELS
+from ...extras.misc import use_modelscope, use_openmind
 from ...extras.packages import is_gradio_available
 from ..common import save_config
-from ..control import can_quantize, can_quantize_to, check_template, get_model_info, list_checkpoints
+from ..control import can_quantize, can_quantize_to, check_template, get_model_info, list_checkpoints, switch_hub
 
 
 if is_gradio_available():
@@ -33,8 +34,10 @@ def create_top() -> dict[str, "Component"]:
     with gr.Row():
         lang = gr.Dropdown(choices=["en", "ru", "zh", "ko", "ja"], value=None, scale=1)
         available_models = list(SUPPORTED_MODELS.keys()) + ["Custom"]
-        model_name = gr.Dropdown(choices=available_models, value=None, scale=3)
-        model_path = gr.Textbox(scale=3)
+        model_name = gr.Dropdown(choices=available_models, value=None, scale=2)
+        model_path = gr.Textbox(scale=2)
+        default_hub = "modelscope" if use_modelscope() else "openmind" if use_openmind() else "huggingface"
+        hub_name = gr.Dropdown(choices=["huggingface", "modelscope", "openmind"], value=default_hub, scale=2)
 
     with gr.Row():
         finetuning_type = gr.Dropdown(choices=METHODS, value="lora", scale=1)
@@ -50,18 +53,25 @@ def create_top() -> dict[str, "Component"]:
     model_name.change(get_model_info, [model_name], [model_path, template], queue=False).then(
         list_checkpoints, [model_name, finetuning_type], [checkpoint_path], queue=False
     ).then(check_template, [lang, template])
-    model_name.input(save_config, inputs=[lang, model_name], queue=False)
-    model_path.input(save_config, inputs=[lang, model_name, model_path], queue=False)
+    model_name.input(save_config, inputs=[lang, hub_name, model_name], queue=False)
+    model_path.input(save_config, inputs=[lang, hub_name, model_name, model_path], queue=False)
     finetuning_type.change(can_quantize, [finetuning_type], [quantization_bit], queue=False).then(
         list_checkpoints, [model_name, finetuning_type], [checkpoint_path], queue=False
     )
     checkpoint_path.focus(list_checkpoints, [model_name, finetuning_type], [checkpoint_path], queue=False)
     quantization_method.change(can_quantize_to, [quantization_method], [quantization_bit], queue=False)
+    hub_name.change(switch_hub, inputs=[hub_name], queue=False).then(
+        get_model_info, [model_name], [model_path, template], queue=False
+    ).then(list_checkpoints, [model_name, finetuning_type], [checkpoint_path], queue=False).then(
+        check_template, [lang, template]
+    )
+    hub_name.input(save_config, inputs=[lang, hub_name], queue=False)
 
     return dict(
         lang=lang,
         model_name=model_name,
         model_path=model_path,
+        hub_name=hub_name,
         finetuning_type=finetuning_type,
         checkpoint_path=checkpoint_path,
         quantization_bit=quantization_bit,

src/llamafactory/webui/control.py

@@ -38,6 +38,15 @@ if is_gradio_available():
     import gradio as gr
 
 
+def switch_hub(hub_name: str) -> None:
+    r"""Switch model hub.
+
+    Inputs: top.hub_name
+    """
+    os.environ["USE_MODELSCOPE_HUB"] = "1" if hub_name == "modelscope" else "0"
+    os.environ["USE_OPENMIND_HUB"] = "1" if hub_name == "openmind" else "0"
+
+
 def can_quantize(finetuning_type: str) -> "gr.Dropdown":
     r"""Judge if the quantization is available in this finetuning type.
 
@@ -112,7 +121,7 @@ def get_trainer_info(lang: str, output_path: os.PathLike, do_train: bool) -> tup
     running_log_path = os.path.join(output_path, RUNNING_LOG)
     if os.path.isfile(running_log_path):
         with open(running_log_path, encoding="utf-8") as f:
-            running_log = f.read()[-20000:]  # avoid lengthy log
+            running_log = "```\n" + f.read()[-20000:] + "\n```\n"  # avoid lengthy log
 
     trainer_log_path = os.path.join(output_path, TRAINER_LOG)
     if os.path.isfile(trainer_log_path):

src/llamafactory/webui/engine.py

@@ -49,11 +49,13 @@ class Engine:
     def resume(self):
         r"""Get the initial value of gradio components and restores training status if necessary."""
         user_config = load_config() if not self.demo_mode else {}  # do not use config in demo mode
-        lang = user_config.get("lang", None) or "en"
+        lang = user_config.get("lang") or "en"
         init_dict = {"top.lang": {"value": lang}, "infer.chat_box": {"visible": self.chatter.loaded}}
 
         if not self.pure_chat:
             current_time = get_time()
+            hub_name = user_config.get("hub_name") or "huggingface"
+            init_dict["top.hub_name"] = {"value": hub_name}
             init_dict["train.current_time"] = {"value": current_time}
             init_dict["train.output_dir"] = {"value": f"train_{current_time}"}
             init_dict["train.config_path"] = {"value": f"{current_time}.yaml"}

src/llamafactory/webui/interface.py

@@ -22,6 +22,7 @@ from .components import (
     create_chat_box,
     create_eval_tab,
     create_export_tab,
+    create_footer,
     create_infer_tab,
     create_top,
     create_train_tab,
@@ -38,15 +39,13 @@ def create_ui(demo_mode: bool = False) -> "gr.Blocks":
     engine = Engine(demo_mode=demo_mode, pure_chat=False)
     hostname = os.getenv("HOSTNAME", os.getenv("COMPUTERNAME", platform.node())).split(".")[0]
 
-    with gr.Blocks(title=f"LLaMA Board ({hostname})", css=CSS) as demo:
+    with gr.Blocks(title=f"LLaMA Factory ({hostname})", css=CSS) as demo:
+        title = gr.HTML()
+        subtitle = gr.HTML()
         if demo_mode:
-            gr.HTML("<h1><center>LLaMA Board: A One-stop Web UI for Getting Started with LLaMA Factory</center></h1>")
-            gr.HTML(
-                '<h3><center>Visit <a href="https://github.com/hiyouga/LLaMA-Factory" target="_blank">'
-                "LLaMA Factory</a> for details.</center></h3>"
-            )
             gr.DuplicateButton(value="Duplicate Space for private use", elem_classes="duplicate-button")
 
+        engine.manager.add_elems("head", {"title": title, "subtitle": subtitle})
         engine.manager.add_elems("top", create_top())
         lang: gr.Dropdown = engine.manager.get_elem_by_id("top.lang")
 
@@ -63,6 +62,7 @@ def create_ui(demo_mode: bool = False) -> "gr.Blocks":
             with gr.Tab("Export"):
                 engine.manager.add_elems("export", create_export_tab(engine))
 
+        engine.manager.add_elems("footer", create_footer())
         demo.load(engine.resume, outputs=engine.manager.get_elem_list(), concurrency_limit=None)
         lang.change(engine.change_lang, [lang], engine.manager.get_elem_list(), queue=False)
         lang.input(save_config, inputs=[lang], queue=False)

src/llamafactory/webui/locales.py

@@ -13,6 +13,55 @@
 # limitations under the License.
 
 LOCALES = {
+    "title": {
+        "en": {
+            "value": "<h1><center>🦙🏭LLaMA Factory: Unified Efficient Fine-Tuning of 100+ LLMs</center></h1>",
+        },
+        "ru": {
+            "value": "<h1><center>🦙🏭LLaMA Factory: Унифицированная эффективная тонкая настройка 100+ LLMs</center></h1>",
+        },
+        "zh": {
+            "value": "<h1><center>🦙🏭LLaMA Factory: 一站式大模型高效微调平台</center></h1>",
+        },
+        "ko": {
+            "value": "<h1><center>🦙🏭LLaMA Factory: 100+ LLMs를 위한 통합 효율적인 튜닝</center></h1>",
+        },
+        "ja": {
+            "value": "<h1><center>🦙🏭LLaMA Factory: 100+ LLMs の統合効率的なチューニング</center></h1>",
+        },
+    },
+    "subtitle": {
+        "en": {
+            "value": (
+                "<h3><center>Visit <a href='https://github.com/hiyouga/LLaMA-Factory' target='_blank'>"
+                "GitHub Page</a></center></h3>"
+            ),
+        },
+        "ru": {
+            "value": (
+                "<h3><center>Посетить <a href='https://github.com/hiyouga/LLaMA-Factory' target='_blank'>"
+                "страницу GitHub</a></center></h3>"
+            ),
+        },
+        "zh": {
+            "value": (
+                "<h3><center>访问 <a href='https://github.com/hiyouga/LLaMA-Factory' target='_blank'>"
+                "GitHub 主页</a></center></h3>"
+            ),
+        },
+        "ko": {
+            "value": (
+                "<h3><center><a href='https://github.com/hiyouga/LLaMA-Factory' target='_blank'>"
+                "GitHub 페이지</a>를 방문하세요.</center></h3>"
+            ),
+        },
+        "ja": {
+            "value": (
+                "<h3><center><a href='https://github.com/hiyouga/LLaMA-Factory' target='_blank'>"
+                "GitHub ページ</a>にアクセスする</center></h3>"
+            ),
+        },
+    },
     "lang": {
         "en": {
             "label": "Language",
@@ -74,6 +123,28 @@ LOCALES = {
             "info": "事前学習済みモデルへのパス、または Hugging Face のモデル識別子。",
         },
     },
+    "hub_name": {
+        "en": {
+            "label": "Hub name",
+            "info": "Choose the model download source.",
+        },
+        "ru": {
+            "label": "Имя хаба",
+            "info": "Выберите источник загрузки модели.",
+        },
+        "zh": {
+            "label": "模型下载源",
+            "info": "选择模型下载源。（网络受限环境推荐使用 ModelScope）",
+        },
+        "ko": {
+            "label": "모델 다운로드 소스",
+            "info": "모델 다운로드 소스를 선택하세요.",
+        },
+        "ja": {
+            "label": "モデルダウンロードソース",
+            "info": "モデルをダウンロードするためのソースを選択してください。",
+        },
+    },
     "finetuning_type": {
         "en": {
             "label": "Finetuning method",
@@ -2849,6 +2920,28 @@ LOCALES = {
             "value": "エクスポート",
         },
     },
+    "device_memory": {
+        "en": {
+            "label": "Device memory",
+            "info": "Current memory usage of the device (GB).",
+        },
+        "ru": {
+            "label": "Память устройства",
+            "info": "Текущая память на устройстве (GB).",
+        },
+        "zh": {
+            "label": "设备显存",
+            "info": "当前设备的显存（GB）。",
+        },
+        "ko": {
+            "label": "디바이스 메모리",
+            "info": "지금 사용 중인 기기 메모리 (GB).",
+        },
+        "ja": {
+            "label": "デバイスメモリ",
+            "info": "現在のデバイスのメモリ（GB）。",
+        },
+    },
 }
 
 

src/llamafactory/webui/runner.py

@@ -16,14 +16,13 @@ import json
 import os
 from collections.abc import Generator
 from copy import deepcopy
-from subprocess import Popen, TimeoutExpired
+from subprocess import PIPE, Popen, TimeoutExpired
 from typing import TYPE_CHECKING, Any, Optional
 
-from transformers.trainer import TRAINING_ARGS_NAME
 from transformers.utils import is_torch_npu_available
 
 from ..extras.constants import LLAMABOARD_CONFIG, MULTIMODAL_SUPPORTED_MODELS, PEFT_METHODS, TRAINING_STAGES
-from ..extras.misc import is_accelerator_available, torch_gc, use_ray
+from ..extras.misc import is_accelerator_available, torch_gc
 from ..extras.packages import is_gradio_available
 from .common import (
     DEFAULT_CACHE_DIR,
@@ -114,7 +113,7 @@ class Runner:
 
         return ""
 
-    def _finalize(self, lang: str, finish_info: str) -> str:
+    def _finalize(self, lang: str, finish_info: str) -> None:
         r"""Clean the cached memory and resets the runner."""
         finish_info = ALERTS["info_aborted"][lang] if self.aborted else finish_info
         gr.Info(finish_info)
@@ -123,7 +122,6 @@ class Runner:
         self.running = False
         self.running_data = None
         torch_gc()
-        return finish_info
 
     def _parse_train_args(self, data: dict["Component", Any]) -> dict[str, Any]:
         r"""Build and validate the training arguments."""
@@ -314,11 +312,13 @@ class Runner:
             max_samples=int(get("eval.max_samples")),
             per_device_eval_batch_size=get("eval.batch_size"),
             predict_with_generate=True,
+            report_to="none",
             max_new_tokens=get("eval.max_new_tokens"),
             top_p=get("eval.top_p"),
             temperature=get("eval.temperature"),
             output_dir=get_save_dir(model_name, finetuning_type, get("eval.output_dir")),
             trust_remote_code=True,
+            ddp_timeout=180000000,
         )
 
         if get("eval.predict"):
@@ -375,7 +375,7 @@ class Runner:
                 env["FORCE_TORCHRUN"] = "1"
 
             # NOTE: DO NOT USE shell=True to avoid security risk
-            self.trainer = Popen(["llamafactory-cli", "train", save_cmd(args)], env=env)
+            self.trainer = Popen(["llamafactory-cli", "train", save_cmd(args)], env=env, stderr=PIPE, text=True)
             yield from self.monitor()
 
     def _build_config_dict(self, data: dict["Component", Any]) -> dict[str, Any]:
@@ -417,7 +417,8 @@ class Runner:
         swanlab_link = self.manager.get_elem_by_id("train.swanlab_link") if self.do_train else None
 
         running_log = ""
-        while self.trainer is not None:
+        return_code = -1
+        while return_code == -1:
             if self.aborted:
                 yield {
                     output_box: ALERTS["info_aborting"][lang],
@@ -436,27 +437,26 @@ class Runner:
                     return_dict[swanlab_link] = running_info["swanlab_link"]
 
                 yield return_dict
+
             try:
-                self.trainer.wait(2)
-                self.trainer = None
+                stderr = self.trainer.communicate(timeout=2)[1]
+                return_code = self.trainer.returncode
             except TimeoutExpired:
                 continue
 
-        if self.do_train:
-            if os.path.exists(os.path.join(output_path, TRAINING_ARGS_NAME)) or use_ray():
+        if return_code == 0 or self.aborted:
             finish_info = ALERTS["info_finished"][lang]
+            if self.do_train:
+                finish_log = ALERTS["info_finished"][lang] + "\n\n" + running_log
             else:
-                finish_info = ALERTS["err_failed"][lang]
-        else:
-            if os.path.exists(os.path.join(output_path, "all_results.json")) or use_ray():
-                finish_info = load_eval_results(os.path.join(output_path, "all_results.json"))
+                finish_log = load_eval_results(os.path.join(output_path, "all_results.json")) + "\n\n" + running_log
         else:
+            print(stderr)
             finish_info = ALERTS["err_failed"][lang]
+            finish_log = ALERTS["err_failed"][lang] + f" Exit code: {return_code}\n\n```\n{stderr}\n```\n"
 
-        return_dict = {
-            output_box: self._finalize(lang, finish_info) + "\n\n" + running_log,
-            progress_bar: gr.Slider(visible=False),
-        }
+        self._finalize(lang, finish_info)
+        return_dict = {output_box: finish_log, progress_bar: gr.Slider(visible=False)}
         yield return_dict
 
     def save_args(self, data):

tests/data/test_formatter.py

@@ -110,8 +110,8 @@ def test_glm4_function_formatter():
 def test_glm4_tool_formatter():
     formatter = ToolFormatter(tool_format="glm4")
     assert formatter.apply(content=json.dumps(TOOLS)) == [
-        "你是一个名为 ChatGLM 的人工智能助手。你是基于智谱AI训练的语言模型 GLM-4 模型开发的，"
-        "你的任务是针对用户的问题和要求提供适当的答复和支持。# 可用工具\n\n"
+        "你是一个名为 ChatGLM 的人工智能助手。你是基于智谱 AI 公司训练的语言模型 GLM-4 模型开发的，"
+        "你的任务是针对用户的问题和要求提供适当的答复和支持。\n\n# 可用工具\n\n"
         f"## test_tool\n\n{json.dumps(TOOLS[0], indent=4, ensure_ascii=False)}\n"
         "在调用上述函数时，请使用 Json 格式表示调用的参数。"
     ]