Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -1,64 +1,64 @@
|
|
|
|
|
|
1 |
import gradio as gr
|
2 |
-
from
|
|
|
|
|
3 |
|
4 |
-
|
5 |
-
|
6 |
-
""
|
7 |
-
client = InferenceClient("google/gemma-2-2b-it")
|
8 |
|
|
|
|
|
|
|
9 |
|
10 |
-
|
11 |
-
message,
|
12 |
-
history: list[tuple[str, str]],
|
13 |
-
system_message,
|
14 |
-
max_tokens,
|
15 |
-
temperature,
|
16 |
-
top_p,
|
17 |
-
):
|
18 |
-
messages = [{"role": "system", "content": system_message}]
|
19 |
|
20 |
-
|
21 |
-
|
22 |
-
messages.append({"role": "user", "content": val[0]})
|
23 |
-
if val[1]:
|
24 |
-
messages.append({"role": "assistant", "content": val[1]})
|
25 |
|
26 |
-
|
|
|
|
|
27 |
|
28 |
-
|
29 |
|
30 |
-
|
31 |
-
|
32 |
-
max_tokens=max_tokens,
|
33 |
-
stream=True,
|
34 |
-
temperature=temperature,
|
35 |
-
top_p=top_p,
|
36 |
-
):
|
37 |
-
token = message.choices[0].delta.content
|
38 |
|
39 |
-
|
40 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
41 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
|
43 |
-
|
44 |
-
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
|
45 |
-
"""
|
46 |
-
demo = gr.ChatInterface(
|
47 |
-
respond,
|
48 |
-
additional_inputs=[
|
49 |
-
gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
|
50 |
-
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
|
51 |
-
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
|
52 |
-
gr.Slider(
|
53 |
-
minimum=0.1,
|
54 |
-
maximum=1.0,
|
55 |
-
value=0.95,
|
56 |
-
step=0.05,
|
57 |
-
label="Top-p (nucleus sampling)",
|
58 |
-
),
|
59 |
-
],
|
60 |
-
)
|
61 |
|
|
|
62 |
|
63 |
-
|
64 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import torch
|
3 |
import gradio as gr
|
4 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
5 |
+
from peft import PeftModel
|
6 |
+
from huggingface_hub import login
|
7 |
|
8 |
+
# 🔥 Installation/mise à jour des dépendances uniquement si nécessaire
|
9 |
+
print("🚀 Vérification et mise à jour des dépendances...")
|
10 |
+
os.system("pip install --no-cache-dir -U transformers peft accelerate torch bitsandbytes scipy")
|
|
|
11 |
|
12 |
+
# 🔥 Correction de `libstdc++6` pour éviter les erreurs `bitsandbytes`
|
13 |
+
os.system("apt-get update && apt-get install -y --reinstall libstdc++6")
|
14 |
+
os.system("ln -sf /usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.30 /usr/lib/x86_64-linux-gnu/libstdc++.so.6")
|
15 |
|
16 |
+
print("✅ Dépendances corrigées et mises à jour !")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
|
18 |
+
# 📌 Authentification Hugging Face
|
19 |
+
login(token=os.getenv("HF_TOKEN"))
|
|
|
|
|
|
|
20 |
|
21 |
+
# 📌 Définition des modèles
|
22 |
+
BASE_MODEL = "google/gemma-2-2b-it"
|
23 |
+
LORA_MODEL = "Serveurperso/gemma-2-2b-it-LoRA"
|
24 |
|
25 |
+
print("🚀 Chargement du modèle Gemma 2B avec LoRA Mémé Ginette...")
|
26 |
|
27 |
+
# 📌 Gestion automatique CPU/GPU
|
28 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
|
30 |
+
# 📌 Chargement du modèle principal
|
31 |
+
try:
|
32 |
+
model = AutoModelForCausalLM.from_pretrained(
|
33 |
+
BASE_MODEL,
|
34 |
+
device_map="auto" if torch.cuda.is_available() else "cpu", # Auto sur GPU si dispo
|
35 |
+
torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
|
36 |
+
trust_remote_code=True
|
37 |
+
)
|
38 |
|
39 |
+
# 📌 Application du LoRA
|
40 |
+
model = PeftModel.from_pretrained(
|
41 |
+
model,
|
42 |
+
LORA_MODEL,
|
43 |
+
device_map="auto" if torch.cuda.is_available() else "cpu",
|
44 |
+
torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
|
45 |
+
)
|
46 |
|
47 |
+
tokenizer = AutoTokenizer.from_pretrained(LORA_MODEL)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
|
49 |
+
print("✅ Modèle chargé avec succès !")
|
50 |
|
51 |
+
except Exception as e:
|
52 |
+
print(f"❌ Erreur lors du chargement du modèle: {e}")
|
53 |
+
exit(1) # Stoppe l’exécution en cas de problème
|
54 |
+
|
55 |
+
# 📌 Interface Gradio
|
56 |
+
def chat(message):
|
57 |
+
inputs = tokenizer(message, return_tensors="pt").to(device)
|
58 |
+
outputs = model.generate(**inputs, max_length=128)
|
59 |
+
return tokenizer.decode(outputs[0], skip_special_tokens=True)
|
60 |
+
|
61 |
+
iface = gr.Interface(fn=chat, inputs="text", outputs="text", title="Mémé Ginette Chatbot")
|
62 |
+
|
63 |
+
print("🚀 Interface Gradio lancée sur port 7860")
|
64 |
+
iface.launch(share=True)
|