Spaces:
Sleeping
Sleeping
s-a-malik
commited on
Commit
·
f89d8b2
1
Parent(s):
3067e7b
tidy
Browse files
app.py
CHANGED
@@ -1,242 +1,11 @@
|
|
1 |
-
# from pathlib import Path
|
2 |
-
# from typing import List, Optional, Tuple
|
3 |
-
|
4 |
-
# import spaces
|
5 |
-
# import gradio as gr
|
6 |
-
# import numpy as np
|
7 |
-
# import torch
|
8 |
-
# from sudachipy import dictionary
|
9 |
-
# from sudachipy import tokenizer as sudachi_tokenizer
|
10 |
-
# from transformers import AutoModelForCausalLM, PreTrainedTokenizer, T5Tokenizer
|
11 |
-
|
12 |
-
|
13 |
-
# model_dir = Path(__file__).parents[0] / "model"
|
14 |
-
# device = torch.device("cuda:0") if torch.cuda.is_available() else torch.device("cpu")
|
15 |
-
# tokenizer = T5Tokenizer.from_pretrained(model_dir)
|
16 |
-
# tokenizer.do_lower_case = True
|
17 |
-
# trained_model = AutoModelForCausalLM.from_pretrained(model_dir)
|
18 |
-
# trained_model.to(device)
|
19 |
-
|
20 |
-
# # baseline model
|
21 |
-
# baseline_model = AutoModelForCausalLM.from_pretrained("rinna/japanese-gpt2-medium")
|
22 |
-
# baseline_model.to(device)
|
23 |
-
|
24 |
-
# sudachi_tokenizer_obj = dictionary.Dictionary().create()
|
25 |
-
# mode = sudachi_tokenizer.Tokenizer.SplitMode.C
|
26 |
-
|
27 |
-
|
28 |
-
# def sudachi_tokenize(input_text: str) -> List[str]:
|
29 |
-
# morphemes = sudachi_tokenizer_obj.tokenize(input_text, mode)
|
30 |
-
# return [morpheme.surface() for morpheme in morphemes]
|
31 |
-
|
32 |
-
|
33 |
-
# def calc_offsets(tokens: List[str]) -> List[int]:
|
34 |
-
# offsets = [0]
|
35 |
-
# for token in tokens:
|
36 |
-
# offsets.append(offsets[-1] + len(token))
|
37 |
-
# return offsets
|
38 |
-
|
39 |
-
|
40 |
-
# def distribute_surprisals_to_characters(
|
41 |
-
# tokens2surprisal: List[Tuple[str, float]]
|
42 |
-
# ) -> List[Tuple[str, float]]:
|
43 |
-
# tokens2surprisal_by_character: List[Tuple[str, float]] = []
|
44 |
-
# for token, surprisal in tokens2surprisal:
|
45 |
-
# token_len = len(token)
|
46 |
-
# for character in token:
|
47 |
-
# tokens2surprisal_by_character.append((character, surprisal / token_len))
|
48 |
-
# return tokens2surprisal_by_character
|
49 |
-
|
50 |
-
|
51 |
-
# def calculate_surprisals_by_character(
|
52 |
-
# input_text: str, model: AutoModelForCausalLM, tokenizer: PreTrainedTokenizer
|
53 |
-
# ) -> Tuple[float, List[Tuple[str, float]]]:
|
54 |
-
# input_tokens = [
|
55 |
-
# token.replace("▁", "")
|
56 |
-
# for token in tokenizer.tokenize(input_text)
|
57 |
-
# if token != "▁"
|
58 |
-
# ]
|
59 |
-
# input_ids = tokenizer.encode(
|
60 |
-
# "<s>" + input_text, add_special_tokens=False, return_tensors="pt"
|
61 |
-
# ).to(device)
|
62 |
-
|
63 |
-
# logits = model(input_ids)["logits"].squeeze(0)
|
64 |
-
|
65 |
-
# surprisals = []
|
66 |
-
# for i in range(logits.shape[0] - 1):
|
67 |
-
# if input_ids[0][i + 1] == 9:
|
68 |
-
# continue
|
69 |
-
# logit = logits[i]
|
70 |
-
# prob = torch.softmax(logit, dim=0)
|
71 |
-
# neg_logprob = -torch.log(prob)
|
72 |
-
# surprisals.append(neg_logprob[input_ids[0][i + 1]].item())
|
73 |
-
# mean_surprisal = np.mean(surprisals)
|
74 |
-
|
75 |
-
# tokens2surprisal: List[Tuple[str, float]] = []
|
76 |
-
# for token, surprisal in zip(input_tokens, surprisals):
|
77 |
-
# tokens2surprisal.append((token, surprisal))
|
78 |
-
|
79 |
-
# char2surprisal = distribute_surprisals_to_characters(tokens2surprisal)
|
80 |
-
|
81 |
-
# return mean_surprisal, char2surprisal
|
82 |
-
|
83 |
-
|
84 |
-
# def aggregate_surprisals_by_offset(
|
85 |
-
# char2surprisal: List[Tuple[str, float]], offsets: List[int]
|
86 |
-
# ) -> List[Tuple[str, float]]:
|
87 |
-
# tokens2surprisal = []
|
88 |
-
# for i in range(len(offsets) - 1):
|
89 |
-
# start = offsets[i]
|
90 |
-
# end = offsets[i + 1]
|
91 |
-
# surprisal = sum([surprisal for _, surprisal in char2surprisal[start:end]])
|
92 |
-
# token = "".join([char for char, _ in char2surprisal[start:end]])
|
93 |
-
# tokens2surprisal.append((token, surprisal))
|
94 |
-
|
95 |
-
# return tokens2surprisal
|
96 |
-
|
97 |
-
|
98 |
-
# def highlight_token(token: str, score: float):
|
99 |
-
# if score > 0:
|
100 |
-
# html_color = "#%02X%02X%02X" % (
|
101 |
-
# 255,
|
102 |
-
# int(255 * (1 - score)),
|
103 |
-
# int(255 * (1 - score)),
|
104 |
-
# )
|
105 |
-
# else:
|
106 |
-
# html_color = "#%02X%02X%02X" % (
|
107 |
-
# int(255 * (1 + score)),
|
108 |
-
# 255,
|
109 |
-
# int(255 * (1 + score)),
|
110 |
-
# )
|
111 |
-
# return '<span style="background-color: {}; color: black">{}</span>'.format(
|
112 |
-
# html_color, token
|
113 |
-
# )
|
114 |
-
|
115 |
-
|
116 |
-
# def create_highlighted_text(
|
117 |
-
# label: str,
|
118 |
-
# tokens2scores: List[Tuple[str, float]],
|
119 |
-
# mean_surprisal: Optional[float] = None,
|
120 |
-
# ):
|
121 |
-
# if mean_surprisal is None:
|
122 |
-
# highlighted_text = "<h2><b>" + label + "</b></h2>"
|
123 |
-
# else:
|
124 |
-
# highlighted_text = (
|
125 |
-
# "<h2><b>" + label + f"</b>(サプライザル平均値: {mean_surprisal:.3f})</h2>"
|
126 |
-
# )
|
127 |
-
# for token, score in tokens2scores:
|
128 |
-
# highlighted_text += highlight_token(token, score)
|
129 |
-
# return highlighted_text
|
130 |
-
|
131 |
-
|
132 |
-
# def normalize_surprisals(
|
133 |
-
# tokens2surprisal: List[Tuple[str, float]], log_scale: bool = False
|
134 |
-
# ) -> List[Tuple[str, float]]:
|
135 |
-
# if log_scale:
|
136 |
-
# surprisals = [np.log(surprisal) for _, surprisal in tokens2surprisal]
|
137 |
-
# else:
|
138 |
-
# surprisals = [surprisal for _, surprisal in tokens2surprisal]
|
139 |
-
# min_surprisal = np.min(surprisals)
|
140 |
-
# max_surprisal = np.max(surprisals)
|
141 |
-
# surprisals = [
|
142 |
-
# (surprisal - min_surprisal) / (max_surprisal - min_surprisal)
|
143 |
-
# for surprisal in surprisals
|
144 |
-
# ]
|
145 |
-
# assert min(surprisals) >= 0
|
146 |
-
# assert max(surprisals) <= 1
|
147 |
-
# return [
|
148 |
-
# (token, surprisal)
|
149 |
-
# for (token, _), surprisal in zip(tokens2surprisal, surprisals)
|
150 |
-
# ]
|
151 |
-
|
152 |
-
|
153 |
-
# def calculate_surprisal_diff(
|
154 |
-
# tokens2surprisal: List[Tuple[str, float]],
|
155 |
-
# baseline_tokens2surprisal: List[Tuple[str, float]],
|
156 |
-
# scale: float = 100.0,
|
157 |
-
# ):
|
158 |
-
# diff_tokens2surprisal = [
|
159 |
-
# (token, (surprisal - baseline_surprisal) * 100)
|
160 |
-
# for (token, surprisal), (_, baseline_surprisal) in zip(
|
161 |
-
# tokens2surprisal, baseline_tokens2surprisal
|
162 |
-
# )
|
163 |
-
# ]
|
164 |
-
# return diff_tokens2surprisal
|
165 |
-
|
166 |
-
# @spaces.GPU
|
167 |
-
# def main(input_text: str) -> Tuple[str, str, str]:
|
168 |
-
# mean_surprisal, char2surprisal = calculate_surprisals_by_character(
|
169 |
-
# input_text, trained_model, tokenizer
|
170 |
-
# )
|
171 |
-
# offsets = calc_offsets(sudachi_tokenize(input_text))
|
172 |
-
# tokens2surprisal = aggregate_surprisals_by_offset(char2surprisal, offsets)
|
173 |
-
# tokens2surprisal = normalize_surprisals(tokens2surprisal)
|
174 |
-
|
175 |
-
# highlighted_text = create_highlighted_text(
|
176 |
-
# "学習後モデル", tokens2surprisal, mean_surprisal
|
177 |
-
# )
|
178 |
-
|
179 |
-
# (
|
180 |
-
# baseline_mean_surprisal,
|
181 |
-
# baseline_char2surprisal,
|
182 |
-
# ) = calculate_surprisals_by_character(input_text, baseline_model, tokenizer)
|
183 |
-
# baseline_tokens2surprisal = aggregate_surprisals_by_offset(
|
184 |
-
# baseline_char2surprisal, offsets
|
185 |
-
# )
|
186 |
-
# baseline_tokens2surprisal = normalize_surprisals(baseline_tokens2surprisal)
|
187 |
-
# baseline_highlighted_text = create_highlighted_text(
|
188 |
-
# "学習前モデル", baseline_tokens2surprisal, baseline_mean_surprisal
|
189 |
-
# )
|
190 |
-
|
191 |
-
# diff_tokens2surprisal = calculate_surprisal_diff(
|
192 |
-
# tokens2surprisal, baseline_tokens2surprisal, 100.0
|
193 |
-
# )
|
194 |
-
# diff_highlighted_text = create_highlighted_text(
|
195 |
-
# "学習前後の差分", diff_tokens2surprisal, None
|
196 |
-
# )
|
197 |
-
# return (
|
198 |
-
# baseline_highlighted_text,
|
199 |
-
# highlighted_text,
|
200 |
-
# diff_highlighted_text,
|
201 |
-
# )
|
202 |
-
|
203 |
-
|
204 |
-
# if __name__ == "__main__":
|
205 |
-
# demo = gr.Interface(
|
206 |
-
# fn=main,
|
207 |
-
# title="文章の読みやすさを自動評価するAI",
|
208 |
-
# description="文章を入力すると、読みづらい表現は赤く、読みやすい表現は青くハイライトされて出力されます。",
|
209 |
-
# # show_label=True,
|
210 |
-
# inputs=gr.Textbox(
|
211 |
-
# lines=5,
|
212 |
-
# label="文章",
|
213 |
-
# placeholder="ここに文章を入力してください。",
|
214 |
-
# ),
|
215 |
-
# outputs=[
|
216 |
-
# gr.HTML(label="学習前モデル", show_label=True),
|
217 |
-
# gr.HTML(label="学習後モデル", show_label=True),
|
218 |
-
# gr.HTML(label="学習前後の差分", show_label=True),
|
219 |
-
# ],
|
220 |
-
# examples=[
|
221 |
-
# "太郎が二郎を殴った。",
|
222 |
-
# "太郎が二郎に殴った。",
|
223 |
-
# "サイエンスインパクトラボは、国立研究開発法人科学技術振興機構(JST)の「科学と社会」推進部が行う共創プログラムです。「先端の研究開発を行う研究者」と「社会課題解決に取り組むプレイヤー」が約3ヶ月に渡って共創活動を行います。",
|
224 |
-
# "近年、ニューラル言語モデルが自然言語の統語知識をどれほど有しているかを、容認性判断課題を通して検証する研究が行われてきている。しかし、このような言語モデルの統語的評価を行うためのデータセットは、主に英語を中心とした欧米の諸言語を対象に構築されてきた。本研究では、既存のデータセットの問題点を克服しつつ、このようなデータセットが構築されてこなかった日本語を対象とした初めてのデータセットである JCoLA (JapaneseCorpus of Linguistic Acceptability) を構築した上で、それを用いた言語モデルの統語的評価を行った。",
|
225 |
-
# ],
|
226 |
-
# )
|
227 |
-
|
228 |
-
# demo.launch()
|
229 |
-
|
230 |
-
|
231 |
import os
|
232 |
import pickle as pkl
|
233 |
from pathlib import Path
|
234 |
from threading import Thread
|
235 |
-
from typing import List,
|
236 |
|
237 |
import spaces
|
238 |
import gradio as gr
|
239 |
-
import numpy as np
|
240 |
import torch
|
241 |
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
|
242 |
|
@@ -246,20 +15,19 @@ DEFAULT_MAX_NEW_TOKENS = 1024
|
|
246 |
MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
|
247 |
|
248 |
DESCRIPTION = """\
|
249 |
-
|
250 |
-
|
251 |
-
The highlighted text shows the model's uncertainty in real-time, with more intense yellow indicating higher uncertainty.
|
252 |
"""
|
253 |
|
254 |
if torch.cuda.is_available():
|
255 |
model_id = "meta-llama/Llama-2-7b-chat-hf"
|
256 |
-
# TODO load the full model?
|
257 |
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_8bit=True)
|
258 |
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
259 |
tokenizer.use_default_system_prompt = False
|
260 |
|
261 |
# load the probe data
|
262 |
-
# TODO
|
263 |
with open("./model/20240625-131035_demo.pkl", "rb") as f:
|
264 |
probe_data = pkl.load(f)
|
265 |
# take the NQ open one
|
@@ -326,14 +94,9 @@ def generate(
|
|
326 |
highlighted_text = ""
|
327 |
for i in range(1, len(hidden)):
|
328 |
token_embeddings = torch.stack([generated_token[0, 0, :].cpu() for generated_token in hidden[i]]) # (num_layers, hidden_size)
|
329 |
-
# print(token_embeddings.shape)
|
330 |
-
# probe the model
|
331 |
-
# print(token_embeddings.numpy()[layer_range].shape)
|
332 |
concat_layers = token_embeddings.numpy()[layer_range[0]:layer_range[1]].reshape(-1) # (num_layers * hidden_size)
|
333 |
-
# print(concat_layers.shape)
|
334 |
# pred in range [-1, 1]
|
335 |
probe_pred = probe.predict_proba(concat_layers.reshape(1, -1))[0][1] * 2 - 1 # prob of high SE
|
336 |
-
# print(probe_pred.shape, probe_pred)
|
337 |
# decode one token at a time
|
338 |
output_id = outputs.sequences[0, input_ids.shape[1]+i]
|
339 |
output_word = tokenizer.decode(output_id)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import os
|
2 |
import pickle as pkl
|
3 |
from pathlib import Path
|
4 |
from threading import Thread
|
5 |
+
from typing import List, Tuple, Iterator
|
6 |
|
7 |
import spaces
|
8 |
import gradio as gr
|
|
|
9 |
import torch
|
10 |
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
|
11 |
|
|
|
15 |
MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
|
16 |
|
17 |
DESCRIPTION = """\
|
18 |
+
This Space demonstrates the Llama-2-7b-chat model with a semantic uncertainty probe.
|
19 |
+
The highlighted text shows the model's uncertainty in real-time, with green indicating more certain generations and red indicating higher uncertainty.
|
|
|
20 |
"""
|
21 |
|
22 |
if torch.cuda.is_available():
|
23 |
model_id = "meta-llama/Llama-2-7b-chat-hf"
|
24 |
+
# TODO load the full model not the 8bit one?
|
25 |
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_8bit=True)
|
26 |
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
27 |
tokenizer.use_default_system_prompt = False
|
28 |
|
29 |
# load the probe data
|
30 |
+
# TODO compare accuracy and SE probe in different tabs/sections
|
31 |
with open("./model/20240625-131035_demo.pkl", "rb") as f:
|
32 |
probe_data = pkl.load(f)
|
33 |
# take the NQ open one
|
|
|
94 |
highlighted_text = ""
|
95 |
for i in range(1, len(hidden)):
|
96 |
token_embeddings = torch.stack([generated_token[0, 0, :].cpu() for generated_token in hidden[i]]) # (num_layers, hidden_size)
|
|
|
|
|
|
|
97 |
concat_layers = token_embeddings.numpy()[layer_range[0]:layer_range[1]].reshape(-1) # (num_layers * hidden_size)
|
|
|
98 |
# pred in range [-1, 1]
|
99 |
probe_pred = probe.predict_proba(concat_layers.reshape(1, -1))[0][1] * 2 - 1 # prob of high SE
|
|
|
100 |
# decode one token at a time
|
101 |
output_id = outputs.sequences[0, input_ids.shape[1]+i]
|
102 |
output_word = tokenizer.decode(output_id)
|