Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
@@ -41,6 +41,9 @@ if torch.cuda.is_available():
|
|
41 |
modelB = AutoModelForCausalLM.from_pretrained(modelB_id, torch_dtype=torch.float16, device_map="auto")
|
42 |
tokenizerB = AutoTokenizer.from_pretrained(modelB_id)
|
43 |
tokenizerB.use_default_system_prompt = False
|
|
|
|
|
|
|
44 |
|
45 |
def make_prompt(entry):
|
46 |
return f"### Human: Don't repeat the assesments, limit to 500 words {entry} ### Assistant:"
|
@@ -70,8 +73,6 @@ def generate(
|
|
70 |
if model == "A":
|
71 |
model = modelA
|
72 |
tokenizer = tokenizerA
|
73 |
-
|
74 |
-
|
75 |
else:
|
76 |
model = modelB
|
77 |
tokenizer = tokenizerB
|
|
|
41 |
modelB = AutoModelForCausalLM.from_pretrained(modelB_id, torch_dtype=torch.float16, device_map="auto")
|
42 |
tokenizerB = AutoTokenizer.from_pretrained(modelB_id)
|
43 |
tokenizerB.use_default_system_prompt = False
|
44 |
+
tokenizerB.pad_token = tokenizerB.eos_token
|
45 |
+
|
46 |
+
|
47 |
|
48 |
def make_prompt(entry):
|
49 |
return f"### Human: Don't repeat the assesments, limit to 500 words {entry} ### Assistant:"
|
|
|
73 |
if model == "A":
|
74 |
model = modelA
|
75 |
tokenizer = tokenizerA
|
|
|
|
|
76 |
else:
|
77 |
model = modelB
|
78 |
tokenizer = tokenizerB
|