Skip to content

Commit d5cb6b3

Browse files
committed
templateミスってた
1 parent c980304 commit d5cb6b3

File tree

3 files changed

+5
-179
lines changed

3 files changed

+5
-179
lines changed

src/dataset/prepare.py

-102
This file was deleted.

src/train.py

+5-5
Original file line numberDiff line numberDiff line change
@@ -36,7 +36,7 @@ def simple_template_for_pretrain(input) -> str:
3636
# inputから、2つ以上連続する改行を除去する
3737
input = "\n".join([line for line in input.splitlines() if line.strip() != ""])
3838
template = f"""\
39-
{input}
39+
{input}\
4040
"""
4141
# Remove any leading whitespace characters from each line in the template.
4242
template = "\n".join([line.lstrip() for line in template.splitlines()])
@@ -50,7 +50,7 @@ def simple_template_for_train(input, output) -> str:
5050
<|im_end|>
5151
<|im_start|>assistant
5252
{output}
53-
<|im_end|>
53+
<|im_end|>\
5454
"""
5555
# Remove any leading whitespace characters from each line in the template.
5656
template = "\n".join([line.lstrip() for line in template.splitlines()])
@@ -65,7 +65,7 @@ def hint_template_for_train(hint, question, answer):
6565
<|im_end|>
6666
<|im_start|>assistant
6767
{answer}
68-
<|im_end|>
68+
<|im_end|>\
6969
"""
7070
# Remove any leading whitespace characters from each line in the template.
7171
template = "\n".join([line.lstrip() for line in template.splitlines()])
@@ -80,7 +80,7 @@ def context_template_for_train(context, question, answer):
8080
<|im_end|>
8181
<|im_start|>assistant
8282
{answer}
83-
<|im_end|>
83+
<|im_end|>\
8484
"""
8585
# Remove any leading whitespace characters from each line in the template.
8686
template = "\n".join([line.lstrip() for line in template.splitlines()])
@@ -98,7 +98,7 @@ def context_hint_template_for_train(hint, context, question, answer):
9898
<|im_end|>
9999
<|im_start|>assistant
100100
{answer}
101-
<|im_end|>
101+
<|im_end|>\
102102
"""
103103
# Remove any leading whitespace characters from each line in the template.
104104
template = "\n".join([line.lstrip() for line in template.splitlines()])

src/train/tokenizer.py

-72
This file was deleted.

0 commit comments

Comments
 (0)