Optimize gradio

graphgen/configs/aggregated_config.yaml

@@ -3,12 +3,12 @@ read:
 split:
   chunk_size: 1024 # chunk size for text splitting
   chunk_overlap: 100 # chunk overlap for text splitting
-output_data_type: aggregated # atomic, aggregated, multi_hop, cot
-output_data_format: ChatML # Alpaca, Sharegpt, ChatML
-tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 search: # web search configuration
   enabled: false # whether to enable web search
   search_types: ["google"] # search engine types, support: google, bing, uniprot, wikipedia
+output_data_type: aggregated # atomic, aggregated, multi_hop, cot
+output_data_format: ChatML # Alpaca, Sharegpt, ChatML
+tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 quiz_and_judge_strategy: # quiz and test whether the LLM masters the knowledge points
   enabled: true
   quiz_samples: 2 # number of quiz samples to generate

graphgen/configs/atomic_config.yaml

@@ -3,12 +3,12 @@ read:
 split:
   chunk_size: 1024 # chunk size for text splitting
   chunk_overlap: 100 # chunk overlap for text splitting
-output_data_type: atomic # atomic, aggregated, multi_hop, cot
-output_data_format: Alpaca # Alpaca, Sharegpt, ChatML
-tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 search: # web search configuration
   enabled: false # whether to enable web search
   search_types: ["google"] # search engine types, support: google, bing, uniprot, wikipedia
+output_data_type: atomic # atomic, aggregated, multi_hop, cot
+output_data_format: Alpaca # Alpaca, Sharegpt, ChatML
+tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 quiz_and_judge_strategy: # quiz and test whether the LLM masters the knowledge points
   enabled: true
   quiz_samples: 2 # number of quiz samples to generate

graphgen/configs/cot_config.yaml

@@ -3,12 +3,12 @@ read:
 split:
   chunk_size: 1024 # chunk size for text splitting
   chunk_overlap: 100 # chunk overlap for text splitting
-output_data_type: cot # atomic, aggregated, multi_hop, cot
-output_data_format: Sharegpt # Alpaca, Sharegpt, ChatML
-tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 search: # web search configuration
   enabled: false # whether to enable web search
   search_types: ["google"] # search engine types, support: google, bing, uniprot, wikipedia
+output_data_type: cot # atomic, aggregated, multi_hop, cot
+output_data_format: Sharegpt # Alpaca, Sharegpt, ChatML
+tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 method_params:
   method: leiden
   max_size: 20 # Maximum size of communities

graphgen/configs/multi_hop_config.yaml

@@ -3,12 +3,12 @@ read:
 split:
   chunk_size: 1024 # chunk size for text splitting
   chunk_overlap: 100 # chunk overlap for text splitting
-output_data_type: multi_hop # atomic, aggregated, multi_hop, cot
-output_data_format: ChatML # Alpaca, Sharegpt, ChatML
-tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 search: # web search configuration
   enabled: false # whether to enable web search
   search_types: ["google"] # search engine types, support: google, bing, uniprot, wikipedia
+output_data_type: multi_hop # atomic, aggregated, multi_hop, cot
+output_data_format: ChatML # Alpaca, Sharegpt, ChatML
+tokenizer: cl100k_base # tokenizer for counting tokens, support tiktoken tokenizer names and local tokenizer path
 quiz_and_judge_strategy: # quiz and test whether the LLM masters the knowledge points
   enabled: false
   quiz_samples: 2 # number of quiz samples to generate

webui/app.py

@@ -13,11 +13,10 @@
 from graphgen.models.llm.limitter import RPM, TPM
 from graphgen.utils import set_logger
 from webui.base import WebuiParams
-from webui.cache_utils import cleanup_workspace, setup_workspace
-from webui.count_tokens import count_tokens
 from webui.i18n import Translate
 from webui.i18n import gettext as _
 from webui.test_api import test_api_connection
+from webui.utils import cleanup_workspace, count_tokens, preview_file, setup_workspace
 
 root_dir = files("webui").parent
 sys.path.append(root_dir)
@@ -391,6 +390,58 @@ def sum_tokens(client):
             with gr.Column(scale=1):
                 test_connection_btn = gr.Button(_("Test Connection"))
 
+        with gr.Row(equal_height=True):
+            with gr.Column(scale=1):
+                with gr.Blocks():
+                    with gr.Row(equal_height=True):
+                        with gr.Column(scale=1):
+                            upload_file = gr.File(
+                                label=_("Upload File"),
+                                file_count="single",
+                                file_types=[".txt", ".json", ".jsonl", ".csv"],
+                                interactive=True,
+                            )
+                            examples_dir = os.path.join(root_dir, "webui", "examples")
+                            gr.Examples(
+                                examples=[
+                                    [os.path.join(examples_dir, "txt_demo.txt")],
+                                    [os.path.join(examples_dir, "jsonl_demo.jsonl")],
+                                    [os.path.join(examples_dir, "json_demo.json")],
+                                    [os.path.join(examples_dir, "csv_demo.csv")],
+                                ],
+                                inputs=upload_file,
+                                label=_("Example Files"),
+                                examples_per_page=4,
+                            )
+            with gr.Column(scale=1):
+                with gr.Blocks():
+                    preview_code = gr.Code(
+                        label=_("File Preview"),
+                        interactive=False,
+                        visible=True,
+                        elem_id="preview_code",
+                    )
+                    preview_df = gr.DataFrame(
+                        label=_("File Preview"),
+                        interactive=False,
+                        visible=False,
+                        elem_id="preview_df",
+                    )
+
+        with gr.Blocks():
+            token_counter = gr.DataFrame(
+                label="Token Stats",
+                headers=[
+                    "Source Text Token Count",
+                    "Estimated Token Usage",
+                    "Token Used",
+                ],
+                datatype="str",
+                interactive=False,
+                visible=False,
+                wrap=True,
+            )
+
         with gr.Blocks():
             with gr.Row(equal_height=True):
                 with gr.Column():
@@ -415,46 +466,12 @@ def sum_tokens(client):
                     )
 
         with gr.Blocks():
-            with gr.Row(equal_height=True):
-                with gr.Column(scale=1):
-                    upload_file = gr.File(
-                        label=_("Upload File"),
-                        file_count="single",
-                        file_types=[".txt", ".json", ".jsonl", ".csv"],
-                        interactive=True,
-                    )
-                    examples_dir = os.path.join(root_dir, "webui", "examples")
-                    gr.Examples(
-                        examples=[
-                            [os.path.join(examples_dir, "txt_demo.txt")],
-                            [os.path.join(examples_dir, "jsonl_demo.jsonl")],
-                            [os.path.join(examples_dir, "json_demo.json")],
-                            [os.path.join(examples_dir, "csv_demo.csv")],
-                        ],
-                        inputs=upload_file,
-                        label=_("Example Files"),
-                        examples_per_page=4,
-                    )
-                with gr.Column(scale=1):
-                    output = gr.File(
-                        label="Output(See Github FAQ)",
-                        file_count="single",
-                        interactive=False,
-                    )
-
-        with gr.Blocks():
-            token_counter = gr.DataFrame(
-                label="Token Stats",
-                headers=[
-                    "Source Text Token Count",
-                    "Estimated Token Usage",
-                    "Token Used",
-                ],
-                datatype="str",
-                interactive=False,
-                visible=False,
-                wrap=True,
-            )
+            with gr.Column(scale=1):
+                output = gr.File(
+                    label=_("Output File"),
+                    file_count="single",
+                    interactive=False,
+                )
 
         submit_btn = gr.Button(_("Run GraphGen"))
 
@@ -494,13 +511,13 @@ def sum_tokens(client):
         )
 
         upload_file.change(
-            lambda x: (gr.update(visible=True)),
-            inputs=[upload_file],
-            outputs=[token_counter],
+            preview_file, inputs=upload_file, outputs=[preview_code, preview_df]
+        ).then(
+            lambda x: gr.update(visible=True), inputs=upload_file, outputs=token_counter
         ).then(
             count_tokens,
             inputs=[upload_file, tokenizer, token_counter],
-            outputs=[token_counter],
+            outputs=token_counter,
         )
 
         # run GraphGen

webui/translation.json

@@ -1,8 +1,10 @@
 {
   "en": {
     "Title": "✨Easy-to-use LLM Training Data Generation Framework✨",
+    "\n\n": "\n\n",
     "### [GraphGen](https://github.com/open-sciencelab/GraphGen) ": "### [GraphGen](https://github.com/open-sciencelab/GraphGen) ",
     "Intro": "is a framework for synthetic data generation guided by knowledge graphs, designed to tackle challenges for knowledge-intensive QA generation. \n\nBy uploading your text chunks (such as knowledge in agriculture, healthcare, or marine science) and filling in the LLM API key, you can generate the training data required by **[LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory)** and **[xtuner](https://github.com/InternLM/xtuner)** online. We will automatically delete user information after completion.",
+    "# ": "# ",
     "Use Trainee Model": "Use Trainee Model to identify knowledge blind spots, please keep disable for SiliconCloud",
     "Synthesizer URL Info": "Base URL for the Synthesizer Model API, use SiliconFlow as default",
     "Synthesizer Model Info": "Model for constructing KGs and generating QAs",
@@ -11,16 +13,20 @@
     "SiliconFlow Token for Trainee Model": "SiliconFlow API Key for Trainee Model",
     "Model Config": "Model Configuration",
     "Generation Config": "Generation Config",
+    "API Config": "API Config",
+    "### ": "### ",
     "SiliconFlow Token": "SiliconFlow API Key",
-    "Test Connection": "Test Connection",
     "Upload File": "Upload File",
     "Example Files": "Example Files",
-    "Run GraphGen": "Run GraphGen"
+    "Output File": "Output File",
+    "File Preview": "File Preview"
   },
   "zh": {
     "Title": "✨开箱即用的LLM训练数据生成框架✨",
+    "\n\n": "\n\n",
     "### [GraphGen](https://github.com/open-sciencelab/GraphGen) ": "### [GraphGen](https://github.com/open-sciencelab/GraphGen) ",
     "Intro": "是一个基于知识图谱的数据合成框架，旨在知识密集型任务中生成问答。\n\n 上传你的文本块（如农业、医疗、海洋知识），填写 LLM api key，即可在线生成 **[LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory)**、**[xtuner](https://github.com/InternLM/xtuner)** 所需训练数据。结束后我们将自动删除用户信息。",
+    "# ": "# ",
     "Use Trainee Model": "使用Trainee Model来识别知识盲区，使用硅基流动时请保持禁用",
     "Synthesizer URL Info": "调用合成模型API的URL，默认使用硅基流动",
     "Synthesizer Model Info": "用于构建知识图谱和生成问答的模型",
@@ -29,10 +35,12 @@
     "SiliconFlow Token for Trainee Model": "SiliconFlow Token for Trainee Model",
     "Model Config": "模型配置",
     "Generation Config": "生成配置",
+    "API Config": "API Config",
+    "### ": "### ",
     "SiliconFlow Token": "SiliconFlow Token",
-    "Test Connection": "测试接口",
     "Upload File": "上传文件",
     "Example Files": "示例文件",
-    "Run GraphGen": "运行GraphGen"
+    "Output File": "输出文件",
+    "File Preview": "文件预览"
   }
 }

webui/utils/__init__.py

@@ -0,0 +1,3 @@
+from .cache import cleanup_workspace, setup_workspace
+from .count_tokens import count_tokens
+from .preview_file import preview_file

webui/cache_utils.py → webui/utils/cache.py

@@ -1,6 +1,7 @@
 import os
-import uuid
 import shutil
+import uuid
+
 
 def setup_workspace(folder):
     request_id = str(uuid.uuid4())

webui/utils/preview_file.py

@@ -0,0 +1,29 @@
+import codecs
+import os
+
+import gradio as gr
+import pandas as pd
+
+
+def preview_file(file):
+    if file is None:
+        return gr.update(visible=False), gr.update(visible=False)
+
+    path = file.name
+    ext = os.path.splitext(path)[1].lower()
+
+    try:
+        if ext == ".csv":
+            df = pd.read_csv(path, nrows=10)
+            return gr.update(visible=False), gr.update(value=df, visible=True)
+        with codecs.open(path, "r", encoding="utf-8") as f:
+            text = f.read(5000)
+            if len(text) == 5000:
+                text += "\n\n... (truncated at 5000 chars)"
+        return gr.update(
+            value=text, visible=True, language="json" if ext != ".txt" else None
+        ), gr.update(visible=False)
+    except Exception as e:  # pylint: disable=broad-except
+        return gr.update(
+            value=f"Preview failed: {e}", visible=True, language=None
+        ), gr.update(visible=False)