filipealmeida
commited on
Commit
•
132970b
1
Parent(s):
16cbd47
Use 4bit model
Browse files
app.py
CHANGED
@@ -46,7 +46,7 @@ def generate_text(prompt, example):
|
|
46 |
|
47 |
|
48 |
parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
|
49 |
-
parser.add_argument("--model-filename", help="Path to the model file", default="ggml-model-
|
50 |
parser.add_argument("--model-local", help="Path to the model file")
|
51 |
parser.add_argument("--gpu", help="How many GPU layers to use", default=0, type=int)
|
52 |
parser.add_argument("--ctx", help="How many context tokens to use", default=1024, type=int)
|
|
|
46 |
|
47 |
|
48 |
parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
|
49 |
+
parser.add_argument("--model-filename", help="Path to the model file", default="ggml-model-Q4_0.gguf")
|
50 |
parser.add_argument("--model-local", help="Path to the model file")
|
51 |
parser.add_argument("--gpu", help="How many GPU layers to use", default=0, type=int)
|
52 |
parser.add_argument("--ctx", help="How many context tokens to use", default=1024, type=int)
|