Readme

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/ankira.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,5 +1,5 @@
 ---
-title: German Dictation Trainer
+title: Ankira - German Dictation Trainer
 emoji: ✍️
 colorFrom: indigo
 colorTo: green
@@ -9,12 +9,27 @@ app_file: app.py
 pinned: false
 ---
 
-# German Dictation Trainer
+<p align="center">
+  <img src="assets/ankira.png" alt="Ankira" width="320">
+</p>
+
+# Ankira: German Dictation Trainer
 
 A mobile-first dictation trainer. Enter the words you want to practice, get a
 short German dictation read aloud, write it by hand, then photograph your page
 to see a word-level diff against the original.
 
+Original plan included importing well-remembered words from Anki cards for practice, thus ANKIra.
+
+# Models used
+
+- https://huggingface.co/LiquidAI/LFM2.5-8B-A1B-GGUF
+- https://huggingface.co/Qwen/Qwen3-VL-8B-Instruct-GGUF
+- https://huggingface.co/bosonai/higgs-audio-v3-tts-4b
+
+
+# Running
+
 No models run in this Space — every inference is an HTTP call to a Modal
 endpoint. Configure these as **Space secrets**:
 
@@ -24,5 +39,7 @@ endpoint. Configure these as **Space secrets**:
 | `MODAL_TTS_URL` | Higgs Audio v3 TTS (`/v1/audio/speech`; base URL also accepted) |
 | `MODAL_OCR_URL` | Qwen3-VL llama-server base URL (multimodal chat completions) |
 
-The recognized-text box is shown on purpose: handwriting OCR is imperfect, so
-seeing what was read keeps OCR errors distinguishable from your own.
+# Our Team
+
+- [@nofate](https://huggingface.co/nofater) Michael Gamov
+- [@latticetower](https://huggingface.co/latticetower) Tatiana Malygina

app.py

@@ -11,6 +11,7 @@ Run locally from this directory (the Space root):
   MODAL_LLM_URL=... MODAL_TTS_URL=... MODAL_OCR_URL=... uv run python app.py
 """
 
+import base64
 import json
 import os
 import tempfile
@@ -101,11 +102,21 @@ MOBILE_CSS = """
     margin: 0 auto !important;
     padding: 12px 14px 28px !important;
 }
-.gradio-container h1 {
+.app-header {
+    display: flex;
+    align-items: center;
+    justify-content: center;
+    gap: 12px;
+    margin: 12px 0 4px;
+}
+.app-logo {
+    height: 128px;
+    width: auto;
+    flex: 0 0 auto;
+}
+.app-title {
     font-size: 1.6rem;
     font-weight: 700;
-    text-align: center;
-    margin: 10px 0 2px;
     background: linear-gradient(135deg, #6366f1, #8b5cf6);
     -webkit-background-clip: text;
     background-clip: text;
@@ -146,6 +157,29 @@ FA_HEAD = (
 SPINNER = '<i class="fa-solid fa-spinner fa-spin"></i>'
 
 
+def _logo_data_uri() -> str:
+    """Embed the scaled-down Ankira logo as a base64 PNG data URI, so the header
+    needs no Gradio file-serving allowlist. Empty string if the file is missing
+    (the header then falls back to the ✍️ emoji)."""
+    path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "assets", "ankira_blue.png")
+    try:
+        with open(path, "rb") as f:
+            b64 = base64.b64encode(f.read()).decode("ascii")
+        return f"data:image/png;base64,{b64}"
+    except OSError:
+        return ""
+
+
+_LOGO_URI = _logo_data_uri()
+_LOGO_TAG = (
+    f'<img class="app-logo" alt="Ankira" src="{_LOGO_URI}">' if _LOGO_URI else "✍️"
+)
+HEADER_HTML = (
+    f'<div class="app-header">{_LOGO_TAG}'
+    f'<span class="app-title">German Dictation Trainer</span></div>'
+)
+
+
 def _require_env(name: str) -> str:
     val = os.environ.get(name)
     if not val:
@@ -307,23 +341,24 @@ def build_ui() -> gr.Blocks:
         # localStorage-backed: survives a tab reload while the learner writes (spec §3).
         state = gr.BrowserState({"diktat": "", "created_at": 0})
 
-        gr.Markdown("# ✍️ German Dictation Trainer")
+        gr.HTML(HEADER_HTML)
 
         # Four stacked views; nav() keeps exactly one visible. Order must match
         # wizard.VIEWS.
         with gr.Group(visible=True) as view_input:
-            gr.Markdown("### 1 · Words to practice", elem_classes="panel-title")
+            gr.Markdown("### Let's start!", elem_classes="panel-title")
+
             words_in = gr.Textbox(
                 label="Words to practice",
                 placeholder="Comma- or newline-separated, e.g. angeblich, ablehnen, Apfel",
                 lines=4,
             )
             level_in = gr.Dropdown(["A1", "A2", "B1", "B2"], value="A2", label="Level")
-            start_btn = gr.Button("Start", variant="primary")
+            start_btn = gr.Button("Generate", variant="primary")
             input_status = gr.Markdown(visible=False, sanitize_html=False, elem_classes="status")
 
         with gr.Group(visible=False) as view_listen:
-            gr.Markdown("### 2 · Listen", elem_classes="panel-title")
+            gr.Markdown("### Listen", elem_classes="panel-title")
             audio_out = gr.Audio(
                 type="filepath", interactive=False, label="Dictation audio"
             )
@@ -335,10 +370,10 @@ def build_ui() -> gr.Blocks:
                 finished_btn = gr.Button("Finished", variant="primary")
 
         with gr.Group(visible=False) as view_upload:
-            gr.Markdown("### 3 · Upload your handwriting", elem_classes="panel-title")
+            gr.Markdown("### Upload", elem_classes="panel-title")
             image_in = gr.Image(
                 type="filepath",
-                sources=["upload", "webcam"],
+                sources=["upload", "webcam", "clipboard"],
                 label="Photo of your handwriting",
             )
             with gr.Row():
@@ -347,7 +382,7 @@ def build_ui() -> gr.Blocks:
             upload_status = gr.Markdown(visible=False, sanitize_html=False, elem_classes="status")
 
         with gr.Group(visible=False) as view_results:
-            gr.Markdown("### 4 · Results", elem_classes="panel-title")
+            gr.Markdown("### Results", elem_classes="panel-title")
             recognized_out = gr.Textbox(
                 label="Recognized text (OCR)", interactive=False, lines=4
             )