update cache
Browse files
app.py
CHANGED
|
@@ -69,16 +69,13 @@ async def start_chat():
|
|
| 69 |
model_id,
|
| 70 |
quantization_config=bnb_config,
|
| 71 |
device_map="auto",
|
| 72 |
-
cache_dir=os.path.join(os.getcwd(), ".cache"),
|
| 73 |
)
|
| 74 |
|
| 75 |
# Move model to GPU if available
|
| 76 |
if torch.cuda.is_available():
|
| 77 |
model = model.to("cuda")
|
| 78 |
|
| 79 |
-
tokenizer = AutoTokenizer.from_pretrained(
|
| 80 |
-
model_id, cache_dir=os.path.join(os.getcwd(), ".cache")
|
| 81 |
-
)
|
| 82 |
|
| 83 |
tokenizer.pad_token = tokenizer.eos_token
|
| 84 |
tokenizer.padding_side = "right"
|
|
|
|
| 69 |
model_id,
|
| 70 |
quantization_config=bnb_config,
|
| 71 |
device_map="auto",
|
|
|
|
| 72 |
)
|
| 73 |
|
| 74 |
# Move model to GPU if available
|
| 75 |
if torch.cuda.is_available():
|
| 76 |
model = model.to("cuda")
|
| 77 |
|
| 78 |
+
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
|
|
|
|
|
|
| 79 |
|
| 80 |
tokenizer.pad_token = tokenizer.eos_token
|
| 81 |
tokenizer.padding_side = "right"
|