Spaces:
Sleeping
Sleeping
Qwen/Qwen2.5-7B-Instruct model
Browse files
app.py
CHANGED
|
@@ -35,7 +35,7 @@ with st.sidebar:
|
|
| 35 |
query_mode = st.selectbox("Query Mode", ["hybrid", "naive", "local", "global"], index=0)
|
| 36 |
|
| 37 |
st.divider()
|
| 38 |
-
st.info("Model:
|
| 39 |
st.success("Optimized for 16GB Memory Limit")
|
| 40 |
|
| 41 |
# --- Model Loading ---
|
|
@@ -49,8 +49,9 @@ def load_models():
|
|
| 49 |
st.info("🧠 Loading Embedding Model (CPU)...")
|
| 50 |
embedding_model = SentenceTransformer("all-MiniLM-L6-v2", device="cpu")
|
| 51 |
|
| 52 |
-
st.info("⏳ Loading LLM (
|
| 53 |
-
model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
|
|
|
|
| 54 |
|
| 55 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
| 56 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
|
|
|
| 35 |
query_mode = st.selectbox("Query Mode", ["hybrid", "naive", "local", "global"], index=0)
|
| 36 |
|
| 37 |
st.divider()
|
| 38 |
+
st.info("Model: Qwen/Qwen2.5-7B-Instruct")
|
| 39 |
st.success("Optimized for 16GB Memory Limit")
|
| 40 |
|
| 41 |
# --- Model Loading ---
|
|
|
|
| 49 |
st.info("🧠 Loading Embedding Model (CPU)...")
|
| 50 |
embedding_model = SentenceTransformer("all-MiniLM-L6-v2", device="cpu")
|
| 51 |
|
| 52 |
+
st.info("⏳ Loading LLM (Qwen/Qwen2.5-7B-Instruct...")
|
| 53 |
+
#model_name = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
|
| 54 |
+
model_name = "Qwen/Qwen2.5-7B-Instruct"
|
| 55 |
|
| 56 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
| 57 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|