Upload 2 files

README.md

@@ -8,12 +8,13 @@ sdk_version: "5.29.0"
 app_file: app.py
 pinned: true
 license: mit
+hardware: zero-gpu
 ---
 
 # 🧠 Fénix Brain
 
 LLM de código para La Forja de Fénix.  
-Modelo: **Qwen2.5-Coder-14B-Instruct** · GPU: ZeroGPU
+Modelo: **Qwen2.5-Coder-7B-Instruct** · GPU: ZeroGPU
 
 ## API
 

app.py

@@ -9,15 +9,15 @@ import gradio as gr
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import torch
 
-MODEL_ID = "Qwen/Qwen2.5-Coder-14B-Instruct"
+MODEL_ID = "Qwen/Qwen2.5-Coder-7B-Instruct"
 API_KEY  = os.environ.get("BRAIN_API_KEY", "")
 
 # ── Cargar modelo al arranque (requerido por ZeroGPU) ─────────────────────────
 tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, trust_remote_code=True)
 model     = AutoModelForCausalLM.from_pretrained(
     MODEL_ID,
-    torch_dtype=torch.bfloat16,
-    device_map="cpu",   # ZeroGPU mueve a GPU solo cuando entra al decorador
+    dtype=torch.bfloat16,
+    device_map="cpu",
     trust_remote_code=True,
 )