Upload app.py with huggingface_hub

app.py

@@ -1,8 +1,6 @@
 """
 Hugging Face Space — Vietnamese Sign Language Recognition
 Upload file .npz -> nhận diện từ ký hiệu -> dịch sang câu tiếng Việt
-
-Gradio UI chạy trên HF Spaces (CPU).
 """
 import os
 import sys
@@ -13,16 +11,13 @@ import gradio as gr
 from huggingface_hub import hf_hub_download, snapshot_download
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 
-# ── Thay bằng repo của bạn ─────────────────────────────────────
 MODEL_REPO = "huy00001/vsl-recognition"
-# ──────────────────────────────────────────────────────────────
 
-# Thêm model code vào path
 sys.path.insert(0, os.path.join(os.path.dirname(__file__), "model"))
 from stgcn_ctc import STGCNCTCModel
 
 DEVICE = "cpu"
-_pipeline = None  # lazy load
+_pipeline = None
 
 
 def load_pipeline():
@@ -30,45 +25,33 @@ def load_pipeline():
     if _pipeline is not None:
         return _pipeline
 
-    print("Downloading model files from HF Hub...")
-
-    # Download label_map.json
+    # label_map
     label_map_path = hf_hub_download(repo_id=MODEL_REPO, filename="label_map.json")
     with open(label_map_path, "r", encoding="utf-8") as f:
         data = json.load(f)
-    label2idx = data["label2idx"]
-    classes   = data["classes"]
+    classes     = data["classes"]
     num_classes = len(classes)
 
-    # Download CTC checkpoint
-    ctc_path = hf_hub_download(repo_id=MODEL_REPO, filename="checkpoints/best_ctc.pt")
+    # CTC model
+    ctc_path  = hf_hub_download(repo_id=MODEL_REPO, filename="checkpoints/best_ctc.pt")
     ctc_model = STGCNCTCModel(num_classes=num_classes)
-    ckpt = torch.load(ctc_path, map_location=DEVICE)
+    ckpt      = torch.load(ctc_path, map_location=DEVICE)
     ctc_model.load_state_dict(ckpt["model_state"])
     ctc_model.eval()
 
-    # Download seq2seq model
-    seq2seq_dir = snapshot_download(
-        repo_id=MODEL_REPO,
-        allow_patterns="checkpoints/best_seq2seq/*",
-    )
+    # seq2seq model
+    seq2seq_dir  = snapshot_download(repo_id=MODEL_REPO, allow_patterns="checkpoints/best_seq2seq/*")
     seq2seq_path = os.path.join(seq2seq_dir, "checkpoints", "best_seq2seq")
-    tokenizer = AutoTokenizer.from_pretrained(seq2seq_path)
-    seq2seq   = AutoModelForSeq2SeqLM.from_pretrained(seq2seq_path)
+    tokenizer    = AutoTokenizer.from_pretrained(seq2seq_path)
+    seq2seq      = AutoModelForSeq2SeqLM.from_pretrained(seq2seq_path)
     seq2seq.eval()
 
-    _pipeline = {
-        "ctc":        ctc_model,
-        "tokenizer":  tokenizer,
-        "seq2seq":    seq2seq,
-        "classes":    classes,
-        "num_classes": num_classes,
-    }
-    print("Pipeline loaded.")
+    _pipeline = {"ctc": ctc_model, "tokenizer": tokenizer,
+                 "seq2seq": seq2seq, "classes": classes, "num_classes": num_classes}
     return _pipeline
 
 
-def greedy_decode(log_probs: torch.Tensor, blank: int = 0) -> list[int]:
+def greedy_decode(log_probs, blank=0):
     indices = log_probs.argmax(dim=-1).tolist()
     decoded, prev = [], blank
     for idx in indices:
@@ -79,58 +62,42 @@ def greedy_decode(log_probs: torch.Tensor, blank: int = 0) -> list[int]:
 
 
 def predict(npz_file):
-    """Gradio handler: nhận file .npz, trả về (words, sentence)"""
     if npz_file is None:
         return "Chưa có file", ""
-
     try:
-        p = load_pipeline()
-
-        # Stage 1b — CTC
+        p   = load_pipeline()
         seq = np.load(npz_file, allow_pickle=True)["sequence"].astype(np.float32)
         x   = torch.from_numpy(seq).unsqueeze(0)
+
         with torch.no_grad():
             log_probs, _ = p["ctc"](x)
             log_probs    = log_probs[:, 0, :]
+
         indices = greedy_decode(log_probs)
         words   = [p["classes"][i - 1] for i in indices if 1 <= i <= p["num_classes"]]
 
         if not words:
             return "Không nhận diện được từ nào", ""
 
-        # Stage 2 — seq2seq
-        text = " | ".join(words)
-        enc  = p["tokenizer"](text, return_tensors="pt", truncation=True, max_length=64)
+        text    = " | ".join(words)
+        enc     = p["tokenizer"](text, return_tensors="pt", truncation=True, max_length=64)
         with torch.no_grad():
-            out_ids = p["seq2seq"].generate(**enc, max_new_tokens=128, num_beams=4)
+            out_ids  = p["seq2seq"].generate(**enc, max_new_tokens=128, num_beams=4)
         sentence = p["tokenizer"].decode(out_ids[0], skip_special_tokens=True)
 
-        return " | ".join(words), sentence
-
+        return text, sentence
     except Exception as e:
         return f"Lỗi: {str(e)}", ""
 
 
-# ── Gradio UI ─────────────────────────────────────────────────��
 with gr.Blocks(title="Nhận diện ngôn ngữ ký hiệu tiếng Việt") as demo:
-    gr.Markdown("## Nhận diện ngôn ngữ ký hiệu tiếng Việt")
+    gr.Markdown("## 🤟 Nhận diện ngôn ngữ ký hiệu tiếng Việt")
     gr.Markdown("Upload file `.npz` chứa skeleton sequence để nhận diện.")
-
-    with gr.Row():
-        file_input = gr.File(label="File .npz", file_types=[".npz"], type="filepath")
-
-    btn = gr.Button("Nhận diện", variant="primary")
-
-    with gr.Row():
-        words_out    = gr.Textbox(label="Từ ký hiệu nhận diện được")
-        sentence_out = gr.Textbox(label="Câu tiếng Việt")
-
+    file_input   = gr.File(label="File .npz", file_types=[".npz"])
+    btn          = gr.Button("Nhận diện", variant="primary")
+    words_out    = gr.Textbox(label="Từ ký hiệu nhận diện được")
+    sentence_out = gr.Textbox(label="Câu tiếng Việt")
     btn.click(fn=predict, inputs=file_input, outputs=[words_out, sentence_out])
 
-    gr.Examples(
-        examples=[],  # thêm file .npz mẫu vào đây nếu muốn
-        inputs=file_input,
-    )
-
 if __name__ == "__main__":
     demo.launch()