feat: Add percentage progress to downloading status in UI and extension

firefox-extension/content.js

@@ -56,8 +56,10 @@ browser.runtime.onMessage.addListener(async (request, sender, sendResponse) => {
             const statusRes = await fetch(`${apiUrl}/status`);
             if (statusRes.ok) {
               const statusData = await statusRes.json();
-              if (statusData[modelVersion] === "downloading") {
-                showToast(`Downloading Gemma 4 (${modelVersion})... this may take a few minutes.`, 0);
+              if (statusData[modelVersion] && statusData[modelVersion].startsWith("downloading")) {
+                const parts = statusData[modelVersion].split(": ");
+                const progress = parts.length > 1 ? parts[1] : "...";
+                showToast(`Downloading Gemma 4 (${modelVersion}) ${progress}... this may take a few minutes.`, 0);
               }
             }
           } catch (e) {
@@ -188,8 +190,10 @@ async function processEntirePage() {
         const statusRes = await fetch(`${apiUrl}/status`);
         if (statusRes.ok) {
           const statusData = await statusRes.json();
-          if (statusData[modelVersion] === "downloading") {
-            showToast(`Downloading Gemma 4 (${modelVersion})... this may take a few minutes.`, 0);
+          if (statusData[modelVersion] && statusData[modelVersion].startsWith("downloading")) {
+            const parts = statusData[modelVersion].split(": ");
+            const progress = parts.length > 1 ? parts[1] : "...";
+            showToast(`Downloading Gemma 4 (${modelVersion}) ${progress}... this may take a few minutes.`, 0);
           }
         }
       } catch (e) {}
@@ -203,8 +207,13 @@ async function processEntirePage() {
     
     // Only show analyzing text if not downloading
     const statusResTemp = await fetch(`${apiUrl}/status`).catch(() => null);
-    if (!statusResTemp || !statusResTemp.ok || (await statusResTemp.json())[modelVersion] !== "downloading") {
+    if (!statusResTemp || !statusResTemp.ok) {
        showToast(`FlowRead analyzing page (${processedCount}/${nodesToProcess.length} blocks)...`, 0);
+    } else {
+       const statusJson = await statusResTemp.json();
+       if (!statusJson[modelVersion] || !statusJson[modelVersion].startsWith("downloading")) {
+          showToast(`FlowRead analyzing page (${processedCount}/${nodesToProcess.length} blocks)...`, 0);
+       }
     }
 
     await Promise.all(batch.map(async (node) => {
@@ -287,8 +296,10 @@ async function updateExisting(newSettings) {
                const statusRes = await fetch(`${apiUrl}/status`);
                if (statusRes.ok) {
                  const statusData = await statusRes.json();
-                 if (statusData[modelVersion] === "downloading") {
-                   showToast(`Downloading Gemma 4 (${modelVersion})... this may take a few minutes.`, 0);
+                 if (statusData[modelVersion] && statusData[modelVersion].startsWith("downloading")) {
+                   const parts = statusData[modelVersion].split(": ");
+                   const progress = parts.length > 1 ? parts[1] : "...";
+                   showToast(`Downloading Gemma 4 (${modelVersion}) ${progress}... this may take a few minutes.`, 0);
                  }
                }
              } catch (e) {}

main.py

@@ -233,6 +233,30 @@ def get_study_stats():
         "preferences": preferences
     }
 
+import sys
+import re
+
+class StderrProgressInterceptor:
+    def __init__(self, original):
+        self.original = original
+        self.current_progress = ""
+        self.active_model = None
+
+    def write(self, s):
+        self.original.write(s)
+        match = re.search(r'(\d+)%\|', s)
+        if match and self.active_model:
+            pct = match.group(1)
+            self.current_progress = f"{pct}%"
+            # Update the global status explicitly so the API returns it immediately
+            model_status[self.active_model] = f"downloading: {self.current_progress}"
+            
+    def flush(self):
+        self.original.flush()
+
+stderr_interceptor = StderrProgressInterceptor(sys.stderr)
+sys.stderr = stderr_interceptor
+
 # --- Saliency API (Existing) ---
 models = {}
 tokenizers = {}
@@ -249,7 +273,8 @@ def load_model(model_name: str):
         return models[model_name], tokenizers[model_name]
         
     print(f"Loading {model_name} on {device}...")
-    model_status[model_name] = "downloading"
+    model_status[model_name] = "downloading: 0%"
+    stderr_interceptor.active_model = model_name
     try:
         if model_name == "27b-4a":
             # Use Gemma 4 27B in 4-bit (requires CUDA)
@@ -285,10 +310,12 @@ def load_model(model_name: str):
         models[model_name] = model
         tokenizers[model_name] = tokenizer
         model_status[model_name] = "loaded"
+        stderr_interceptor.active_model = None
         return model, tokenizer
     except Exception as e:
         print(f"Error loading model {model_name}: {e}")
         model_status[model_name] = "error"
+        stderr_interceptor.active_model = None
         raise e
 
 # Pre-load default 2b

static/index.html

@@ -485,8 +485,10 @@
                         const statusRes = await fetch('/status');
                         if (statusRes.ok) {
                             const statusData = await statusRes.json();
-                            if (statusData[modelVersion] === "downloading") {
-                                loading.textContent = `Downloading Gemma 4 (${modelVersion}) Model... this takes a few minutes.`;
+                            if (statusData[modelVersion] && statusData[modelVersion].startsWith("downloading")) {
+                                const parts = statusData[modelVersion].split(": ");
+                                const progress = parts.length > 1 ? parts[1] : "...";
+                                loading.textContent = `Downloading Gemma 4 (${modelVersion}) Model ${progress}... this takes a few minutes.`;
                             }
                         }
                     } catch(e) {}

test_stderr.py

@@ -0,0 +1,21 @@
+import sys
+import re
+
+class StderrInterceptor:
+    def __init__(self, original):
+        self.original = original
+
+    def write(self, s):
+        self.original.write(s)
+        match = re.search(r'(\d+)%\|', s)
+        if match:
+            self.original.write(f"\nINTERCEPTED: {match.group(1)}%\n")
+            self.original.flush()
+    
+    def flush(self):
+        self.original.flush()
+
+sys.stderr = StderrInterceptor(sys.stderr)
+
+from transformers import AutoConfig
+AutoConfig.from_pretrained("gpt2", force_download=True)

test_tqdm.py

@@ -0,0 +1,21 @@
+import huggingface_hub.utils as hf_utils
+
+original_tqdm = hf_utils.tqdm
+
+class MyTqdm(original_tqdm):
+    def update(self, n=1):
+        super().update(n)
+        if hasattr(self, 'total') and self.total:
+            print(f"DL_PROGRESS: {self.n / self.total:.2%}", flush=True)
+
+hf_utils.tqdm = MyTqdm
+hf_utils.tqdm.tqdm = MyTqdm # Also overwrite the module attribute just in case
+
+import huggingface_hub.file_download
+huggingface_hub.file_download.tqdm = MyTqdm
+
+try:
+    from transformers import AutoConfig
+    AutoConfig.from_pretrained("gpt2", force_download=True)
+except Exception as e:
+    print(e)