Spaces:
Runtime error
Runtime error
MohamedRashad
commited on
Commit
•
d2e2dfe
1
Parent(s):
0c0efc4
refactor: improve tokenization for Arabic text
Browse files- app.py +121 -59
- arabic_tokenizers_leaderboard.jsonl +14 -0
app.py
CHANGED
@@ -7,39 +7,74 @@ import random
|
|
7 |
from pathlib import Path
|
8 |
|
9 |
initial_list_of_models = [
|
|
|
10 |
"Xenova/gpt-4o",
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
"NousResearch/Meta-Llama-3-8B",
|
12 |
"CohereForAI/c4ai-command-r-v01",
|
13 |
"CohereForAI/c4ai-command-r-plus",
|
14 |
"core42/jais-13b",
|
|
|
15 |
]
|
16 |
|
17 |
-
dataset = load_dataset("MohamedRashad/rasaif-translations", split="train")["arabic"]
|
18 |
-
|
19 |
dataframe_path = Path(__file__).parent / "arabic_tokenizers_leaderboard.jsonl"
|
20 |
-
|
21 |
if dataframe_path.exists():
|
22 |
df = pd.read_json(dataframe_path, lines=True)
|
23 |
else:
|
24 |
-
df = pd.DataFrame(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
|
26 |
-
for
|
27 |
-
|
28 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
tokenizer = AutoTokenizer.from_pretrained(
|
30 |
model_name, use_fast=True, trust_remote_code=True
|
31 |
)
|
32 |
vocab_size = tokenizer.vocab_size
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
43 |
|
44 |
# Sort the dataframe by the number of tokens
|
45 |
df = df.sort_values(by="➕ Total Number of Tokens", ascending=True)
|
@@ -47,59 +82,57 @@ df = df.sort_values(by="➕ Total Number of Tokens", ascending=True)
|
|
47 |
# Save the dataframe to a csv file
|
48 |
df.to_json(dataframe_path, lines=True, orient="records", force_ascii=False)
|
49 |
|
|
|
50 |
def submit(model_name):
|
51 |
global df
|
52 |
if model_name in df["📛 Models"].values:
|
53 |
-
return
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
df = df._append(
|
60 |
-
{
|
61 |
-
"📛 Models": model_name,
|
62 |
-
"➕ Total Number of Tokens": number_of_tokens,
|
63 |
-
"📘 Vocab Size": vocab_size,
|
64 |
-
"Tokenizer Class": tokenizer.__class__.__name__,
|
65 |
-
},
|
66 |
-
ignore_index=True,
|
67 |
-
)
|
68 |
df = df.sort_values(by="➕ Total Number of Tokens", ascending=True)
|
69 |
df.to_json(dataframe_path, lines=True, orient="records", force_ascii=False)
|
70 |
-
return
|
|
|
|
|
|
|
|
|
|
|
71 |
|
72 |
def generate_distinct_colors(n):
|
73 |
"""Generate n visually distinct colors in hexadecimal format."""
|
74 |
if n > 256**3:
|
75 |
raise ValueError("Cannot generate more than 16,777,216 unique colors.")
|
76 |
-
|
77 |
# To ensure colors are distinct, calculate an appropriate distance between colors
|
78 |
# The cube root of 256**3 (total colors) divided by n gives a crude initial spacing estimate
|
79 |
-
spacing = int((256 * 256 * 256)**(1/3) / n**(1/3))
|
80 |
max_val = 256 - spacing
|
81 |
-
|
82 |
# Set to keep track of used colors
|
83 |
used_colors = set()
|
84 |
-
|
85 |
# List to store the result colors
|
86 |
result = []
|
87 |
-
|
88 |
attempts = 0
|
89 |
while len(result) < n:
|
90 |
# Generate a color with a random start and controlled spacing
|
91 |
r = random.randint(0, max_val)
|
92 |
g = random.randint(0, max_val)
|
93 |
b = random.randint(0, max_val)
|
94 |
-
|
95 |
# Scale up by spacing to ensure minimum distance between colors
|
96 |
r = min(255, r * spacing)
|
97 |
g = min(255, g * spacing)
|
98 |
b = min(255, b * spacing)
|
99 |
-
|
100 |
# Format the color in hexadecimal
|
101 |
color = f"#{r:02X}{g:02X}{b:02X}"
|
102 |
-
|
103 |
# Ensure this color hasn't been used
|
104 |
if color not in used_colors:
|
105 |
used_colors.add(color)
|
@@ -111,29 +144,31 @@ def generate_distinct_colors(n):
|
|
111 |
spacing = max(1, spacing - 1)
|
112 |
max_val = 256 - spacing
|
113 |
attempts = 0
|
114 |
-
|
115 |
return result
|
116 |
|
|
|
117 |
def decode_bpe_tokens(tokens):
|
118 |
fixed_tokens = []
|
119 |
for token in tokens:
|
120 |
# Check if the token starts with the special BPE space character 'Ġ'
|
121 |
-
if token.startswith(
|
122 |
# Process the rest of the token
|
123 |
try:
|
124 |
# Decode the rest of the token from UTF-8 bytes understood as Latin-1 characters
|
125 |
-
fixed_token =
|
126 |
except UnicodeDecodeError:
|
127 |
fixed_token = token # Use the original token if decoding fails
|
128 |
else:
|
129 |
try:
|
130 |
# Directly encode and decode without misinterpretation steps
|
131 |
-
fixed_token = token.encode(
|
132 |
except UnicodeDecodeError:
|
133 |
fixed_token = token # Use the original token if decoding fails
|
134 |
fixed_tokens.append(fixed_token)
|
135 |
return fixed_tokens
|
136 |
|
|
|
137 |
def tokenize_text(text, chosen_model, better_tokenization=False):
|
138 |
tokenizer = AutoTokenizer.from_pretrained(chosen_model)
|
139 |
tokenized_text = decode_bpe_tokens(tokenizer.tokenize(text))
|
@@ -144,11 +179,13 @@ def tokenize_text(text, chosen_model, better_tokenization=False):
|
|
144 |
for token in tokenized_text:
|
145 |
correct_tokenized_text = ""
|
146 |
for char in text:
|
147 |
-
correct_tokenized_text += char
|
148 |
-
current_token = decode_bpe_tokens(
|
|
|
|
|
149 |
if current_token[0] == token:
|
150 |
final_tokenized_text.append(correct_tokenized_text)
|
151 |
-
text = text[len(correct_tokenized_text):]
|
152 |
break
|
153 |
else:
|
154 |
final_tokenized_text = tokenized_text
|
@@ -158,19 +195,30 @@ def tokenize_text(text, chosen_model, better_tokenization=False):
|
|
158 |
color_map = {}
|
159 |
for idx, token in enumerate(final_tokenized_text):
|
160 |
output.append((token, str(idx)))
|
161 |
-
color_map[str(idx+1)] = random_colors[idx % len(random_colors)]
|
162 |
|
163 |
return gr.HighlightedText(output, color_map)
|
164 |
|
|
|
165 |
def refresh():
|
166 |
global df
|
167 |
df = pd.read_json(dataframe_path, lines=True)
|
168 |
-
return
|
|
|
|
|
|
|
|
|
169 |
|
170 |
-
leaderboard_description = """The `Total Number of Tokens` in this leaderboard is based on the total number of tokens
|
171 |
-
This dataset was chosen because it represents Arabic Fusha text in a small and consentrated manner.
|
172 |
|
173 |
-
A tokenizer that scores high in this leaderboard will be efficient in parsing Arabic in its different dialects and forms
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
174 |
"""
|
175 |
|
176 |
with gr.Blocks() as demo:
|
@@ -188,7 +236,7 @@ with gr.Blocks() as demo:
|
|
188 |
y_title=" ",
|
189 |
width=1000,
|
190 |
height=400,
|
191 |
-
tooltip=["📘 Vocab Size", "
|
192 |
vertical=False,
|
193 |
x_label_angle=30,
|
194 |
)
|
@@ -196,10 +244,18 @@ with gr.Blocks() as demo:
|
|
196 |
label="Model Name from Hugging Face (e.g. Xenova/gpt-4o)"
|
197 |
)
|
198 |
with gr.Row():
|
199 |
-
submit_new_model_btn = gr.Button(
|
|
|
|
|
200 |
refresh_btn = gr.Button(value="Refresh", variant="secondary", scale=1)
|
201 |
with gr.Tab(label="Try tokenizers"):
|
202 |
-
text = gr.Textbox(
|
|
|
|
|
|
|
|
|
|
|
|
|
203 |
dropdown = gr.Dropdown(
|
204 |
label="Select a model",
|
205 |
choices=df["📛 Models"].tolist(),
|
@@ -207,12 +263,18 @@ with gr.Blocks() as demo:
|
|
207 |
)
|
208 |
with gr.Row():
|
209 |
submit_text_btn = gr.Button(value="Submit", variant="primary", scale=3)
|
210 |
-
checkbox = gr.Checkbox(
|
|
|
|
|
211 |
tokenized_textbox = gr.HighlightedText(label="Tokenized text")
|
212 |
|
213 |
-
submit_new_model_btn.click(
|
|
|
|
|
214 |
refresh_btn.click(refresh, outputs=[dataframe, barplot, dropdown])
|
215 |
-
submit_text_btn.click(
|
|
|
|
|
216 |
|
217 |
|
218 |
demo.launch()
|
|
|
7 |
from pathlib import Path
|
8 |
|
9 |
initial_list_of_models = [
|
10 |
+
"asafaya/bert-base-arabic",
|
11 |
"Xenova/gpt-4o",
|
12 |
+
"FreedomIntelligence/AceGPT-v1.5-13B-Chat",
|
13 |
+
"FreedomIntelligence/AceGPT-13B",
|
14 |
+
"Qwen/Qwen1.5-7B-Chat",
|
15 |
+
"Qwen/Qwen1.5-110B-Chat",
|
16 |
+
"microsoft/Phi-3-mini-128k-instruct",
|
17 |
+
"unsloth/gemma-2b-bnb-4bit",
|
18 |
"NousResearch/Meta-Llama-3-8B",
|
19 |
"CohereForAI/c4ai-command-r-v01",
|
20 |
"CohereForAI/c4ai-command-r-plus",
|
21 |
"core42/jais-13b",
|
22 |
+
"core42/jais-30b-chat-v3",
|
23 |
]
|
24 |
|
|
|
|
|
25 |
dataframe_path = Path(__file__).parent / "arabic_tokenizers_leaderboard.jsonl"
|
|
|
26 |
if dataframe_path.exists():
|
27 |
df = pd.read_json(dataframe_path, lines=True)
|
28 |
else:
|
29 |
+
df = pd.DataFrame(
|
30 |
+
columns=[
|
31 |
+
"👳 Tokenize Tashkeel",
|
32 |
+
"📛 Models",
|
33 |
+
"🪺 Fertility Score",
|
34 |
+
"➕ Total Number of Tokens",
|
35 |
+
"📘 Vocab Size",
|
36 |
+
"Tokenizer Class",
|
37 |
+
]
|
38 |
+
)
|
39 |
|
40 |
+
# Datasets used for calculating the number of tokens
|
41 |
+
arabic_dataset1 = load_dataset("MohamedRashad/rasaif-translations", split="train")["arabic"]
|
42 |
+
arabic_dataset2 = load_dataset("HeshamHaroon/arabic-quotes", split="train")["quote"]
|
43 |
+
arabic_dataset3 = load_dataset("SaiedAlshahrani/Moroccan_Arabic_Wikipedia_20230101_nobots", split="train")["text"]
|
44 |
+
all_data = arabic_dataset1 + arabic_dataset2 + arabic_dataset3
|
45 |
+
print(f"Total number of samples: {len(all_data)}")
|
46 |
+
all_text = " ".join(all_data)
|
47 |
+
all_words = all_text.split()
|
48 |
+
|
49 |
+
def benchmark_tokenizer(model_name) -> float:
|
50 |
+
# Initialize the tokenizer
|
51 |
tokenizer = AutoTokenizer.from_pretrained(
|
52 |
model_name, use_fast=True, trust_remote_code=True
|
53 |
)
|
54 |
vocab_size = tokenizer.vocab_size
|
55 |
+
total_number_of_tokens = len(tokenizer.tokenize(all_text))
|
56 |
+
|
57 |
+
# Check if the tokenizer maintains the tashkeel
|
58 |
+
dummy_text = "السَّلَامُ عَلَيْكُمْ وَرَحْمَةُ اللَّهِ وَبَرَكَاتُهُ"
|
59 |
+
tokenized_text = tokenizer.decode(tokenizer.encode(dummy_text), skip_special_tokens=True)
|
60 |
+
tashkeel_maintainer = "✅" if tokenized_text == dummy_text else "❌"
|
61 |
+
|
62 |
+
return {
|
63 |
+
"👳 Tokenize Tashkeel": tashkeel_maintainer,
|
64 |
+
"📛 Models": model_name,
|
65 |
+
"🪺 Fertility Score": round(total_number_of_tokens / len(all_words), 3),
|
66 |
+
"📘 Vocab Size": vocab_size,
|
67 |
+
"➕ Total Number of Tokens": total_number_of_tokens,
|
68 |
+
"Tokenizer Class": tokenizer.__class__.__name__,
|
69 |
+
}
|
70 |
+
|
71 |
+
|
72 |
+
for model_name in tqdm(initial_list_of_models):
|
73 |
+
if model_name in df["📛 Models"].values:
|
74 |
+
continue
|
75 |
+
|
76 |
+
benchmark_data = benchmark_tokenizer(model_name)
|
77 |
+
df = df._append(benchmark_data, ignore_index=True)
|
78 |
|
79 |
# Sort the dataframe by the number of tokens
|
80 |
df = df.sort_values(by="➕ Total Number of Tokens", ascending=True)
|
|
|
82 |
# Save the dataframe to a csv file
|
83 |
df.to_json(dataframe_path, lines=True, orient="records", force_ascii=False)
|
84 |
|
85 |
+
|
86 |
def submit(model_name):
|
87 |
global df
|
88 |
if model_name in df["📛 Models"].values:
|
89 |
+
return (
|
90 |
+
gr.Dataframe(df),
|
91 |
+
gr.BarPlot(df),
|
92 |
+
gr.Dropdown(choices=df["📛 Models"].tolist()),
|
93 |
+
)
|
94 |
+
benchmark_data = benchmark_tokenizer(model_name)
|
95 |
+
df = df._append(benchmark_data, ignore_index=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
96 |
df = df.sort_values(by="➕ Total Number of Tokens", ascending=True)
|
97 |
df.to_json(dataframe_path, lines=True, orient="records", force_ascii=False)
|
98 |
+
return (
|
99 |
+
gr.Dataframe(df),
|
100 |
+
gr.BarPlot(df),
|
101 |
+
gr.Dropdown(choices=df["📛 Models"].tolist()),
|
102 |
+
)
|
103 |
+
|
104 |
|
105 |
def generate_distinct_colors(n):
|
106 |
"""Generate n visually distinct colors in hexadecimal format."""
|
107 |
if n > 256**3:
|
108 |
raise ValueError("Cannot generate more than 16,777,216 unique colors.")
|
109 |
+
|
110 |
# To ensure colors are distinct, calculate an appropriate distance between colors
|
111 |
# The cube root of 256**3 (total colors) divided by n gives a crude initial spacing estimate
|
112 |
+
spacing = int((256 * 256 * 256) ** (1 / 3) / n ** (1 / 3))
|
113 |
max_val = 256 - spacing
|
114 |
+
|
115 |
# Set to keep track of used colors
|
116 |
used_colors = set()
|
117 |
+
|
118 |
# List to store the result colors
|
119 |
result = []
|
120 |
+
|
121 |
attempts = 0
|
122 |
while len(result) < n:
|
123 |
# Generate a color with a random start and controlled spacing
|
124 |
r = random.randint(0, max_val)
|
125 |
g = random.randint(0, max_val)
|
126 |
b = random.randint(0, max_val)
|
127 |
+
|
128 |
# Scale up by spacing to ensure minimum distance between colors
|
129 |
r = min(255, r * spacing)
|
130 |
g = min(255, g * spacing)
|
131 |
b = min(255, b * spacing)
|
132 |
+
|
133 |
# Format the color in hexadecimal
|
134 |
color = f"#{r:02X}{g:02X}{b:02X}"
|
135 |
+
|
136 |
# Ensure this color hasn't been used
|
137 |
if color not in used_colors:
|
138 |
used_colors.add(color)
|
|
|
144 |
spacing = max(1, spacing - 1)
|
145 |
max_val = 256 - spacing
|
146 |
attempts = 0
|
147 |
+
|
148 |
return result
|
149 |
|
150 |
+
|
151 |
def decode_bpe_tokens(tokens):
|
152 |
fixed_tokens = []
|
153 |
for token in tokens:
|
154 |
# Check if the token starts with the special BPE space character 'Ġ'
|
155 |
+
if token.startswith("Ġ"):
|
156 |
# Process the rest of the token
|
157 |
try:
|
158 |
# Decode the rest of the token from UTF-8 bytes understood as Latin-1 characters
|
159 |
+
fixed_token = " " + token[1:].encode("utf-8").decode("utf-8")
|
160 |
except UnicodeDecodeError:
|
161 |
fixed_token = token # Use the original token if decoding fails
|
162 |
else:
|
163 |
try:
|
164 |
# Directly encode and decode without misinterpretation steps
|
165 |
+
fixed_token = token.encode("utf-8").decode("utf-8")
|
166 |
except UnicodeDecodeError:
|
167 |
fixed_token = token # Use the original token if decoding fails
|
168 |
fixed_tokens.append(fixed_token)
|
169 |
return fixed_tokens
|
170 |
|
171 |
+
|
172 |
def tokenize_text(text, chosen_model, better_tokenization=False):
|
173 |
tokenizer = AutoTokenizer.from_pretrained(chosen_model)
|
174 |
tokenized_text = decode_bpe_tokens(tokenizer.tokenize(text))
|
|
|
179 |
for token in tokenized_text:
|
180 |
correct_tokenized_text = ""
|
181 |
for char in text:
|
182 |
+
correct_tokenized_text += char
|
183 |
+
current_token = decode_bpe_tokens(
|
184 |
+
tokenizer.tokenize(correct_tokenized_text)
|
185 |
+
)
|
186 |
if current_token[0] == token:
|
187 |
final_tokenized_text.append(correct_tokenized_text)
|
188 |
+
text = text[len(correct_tokenized_text) :]
|
189 |
break
|
190 |
else:
|
191 |
final_tokenized_text = tokenized_text
|
|
|
195 |
color_map = {}
|
196 |
for idx, token in enumerate(final_tokenized_text):
|
197 |
output.append((token, str(idx)))
|
198 |
+
color_map[str(idx + 1)] = random_colors[idx % len(random_colors)]
|
199 |
|
200 |
return gr.HighlightedText(output, color_map)
|
201 |
|
202 |
+
|
203 |
def refresh():
|
204 |
global df
|
205 |
df = pd.read_json(dataframe_path, lines=True)
|
206 |
+
return (
|
207 |
+
gr.Dataframe(df),
|
208 |
+
gr.BarPlot(df),
|
209 |
+
gr.Dropdown(choices=df["📛 Models"].tolist()),
|
210 |
+
)
|
211 |
|
212 |
+
leaderboard_description = """The `Total Number of Tokens` in this leaderboard is based on the total number of tokens got from the Arabic section of [rasaif-translations](https://huggingface.co/datasets/MohamedRashad/rasaif-translations) dataset (This dataset was chosen because it represents Arabic Fusha text in a small and concentrated manner).
|
|
|
213 |
|
214 |
+
**A tokenizer that scores high in this leaderboard will be efficient in parsing Arabic in its different dialects and forms.**
|
215 |
+
|
216 |
+
## Updates
|
217 |
+
1. New datasets is added for the evaluation (e.g. [arabic-quotes](https://huggingface.co/datasets/HeshamHaroon/arabic-quotes), [Moroccan_Arabic_Wikipedia_20230101_nobots](https://huggingface.co/datasets/SaiedAlshahrani/Moroccan_Arabic_Wikipedia_20230101_nobots)).
|
218 |
+
1. `Fertility Score` is calculated by dividing the total number of tokens by the total number of words in the dataset (another way to interpret `Total Number of Tokens`).
|
219 |
+
1. `Tokenize Tashkeel` is an indicator of whether the tokenizer maintains the tashkeel when tokenizing or not (`✅` for yes, `❌` for no).
|
220 |
+
1. `Vocab Size` is the total number of tokens in the tokenizer's vocabulary (e.g. `10000` tokens).
|
221 |
+
1. `Tokenizer Class` is the class of the tokenizer (e.g. `BertTokenizer` or `GPT2Tokenizer`)
|
222 |
"""
|
223 |
|
224 |
with gr.Blocks() as demo:
|
|
|
236 |
y_title=" ",
|
237 |
width=1000,
|
238 |
height=400,
|
239 |
+
tooltip=["📘 Vocab Size", "🪺 Fertility Score"],
|
240 |
vertical=False,
|
241 |
x_label_angle=30,
|
242 |
)
|
|
|
244 |
label="Model Name from Hugging Face (e.g. Xenova/gpt-4o)"
|
245 |
)
|
246 |
with gr.Row():
|
247 |
+
submit_new_model_btn = gr.Button(
|
248 |
+
value="Submit New Model", variant="primary", scale=3
|
249 |
+
)
|
250 |
refresh_btn = gr.Button(value="Refresh", variant="secondary", scale=1)
|
251 |
with gr.Tab(label="Try tokenizers"):
|
252 |
+
text = gr.Textbox(
|
253 |
+
label="Enter a text",
|
254 |
+
lines=5,
|
255 |
+
value="السلام عليكم ورحمة الله",
|
256 |
+
rtl=True,
|
257 |
+
text_align="right",
|
258 |
+
)
|
259 |
dropdown = gr.Dropdown(
|
260 |
label="Select a model",
|
261 |
choices=df["📛 Models"].tolist(),
|
|
|
263 |
)
|
264 |
with gr.Row():
|
265 |
submit_text_btn = gr.Button(value="Submit", variant="primary", scale=3)
|
266 |
+
checkbox = gr.Checkbox(
|
267 |
+
label="Better tokenization for Arabic Text", value=False, scale=1
|
268 |
+
)
|
269 |
tokenized_textbox = gr.HighlightedText(label="Tokenized text")
|
270 |
|
271 |
+
submit_new_model_btn.click(
|
272 |
+
submit, model_name, outputs=[dataframe, barplot, dropdown]
|
273 |
+
)
|
274 |
refresh_btn.click(refresh, outputs=[dataframe, barplot, dropdown])
|
275 |
+
submit_text_btn.click(
|
276 |
+
tokenize_text, inputs=[text, dropdown, checkbox], outputs=[tokenized_textbox]
|
277 |
+
)
|
278 |
|
279 |
|
280 |
demo.launch()
|
arabic_tokenizers_leaderboard.jsonl
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{"👳 Tokenize Tashkeel":"❌","📛 Models":"asafaya\/bert-base-arabic","🪺 Fertility Score":1.614,"➕ Total Number of Tokens":1242530,"📘 Vocab Size":32000,"Tokenizer Class":"BertTokenizerFast"}
|
2 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"core42\/jais-13b","🪺 Fertility Score":1.668,"➕ Total Number of Tokens":1284508,"📘 Vocab Size":84992,"Tokenizer Class":"PreTrainedTokenizerFast"}
|
3 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"core42\/jais-30b-chat-v3","🪺 Fertility Score":1.668,"➕ Total Number of Tokens":1284508,"📘 Vocab Size":84992,"Tokenizer Class":"PreTrainedTokenizerFast"}
|
4 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"FreedomIntelligence\/AceGPT-v1.5-13B-Chat","🪺 Fertility Score":1.888,"➕ Total Number of Tokens":1453838,"📘 Vocab Size":44800,"Tokenizer Class":"LlamaTokenizerFast"}
|
5 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"Xenova\/gpt-4o","🪺 Fertility Score":2.115,"➕ Total Number of Tokens":1628374,"📘 Vocab Size":200000,"Tokenizer Class":"GPT2TokenizerFast"}
|
6 |
+
{"👳 Tokenize Tashkeel":"❌","📛 Models":"CohereForAI\/c4ai-command-r-v01","🪺 Fertility Score":2.154,"➕ Total Number of Tokens":1658463,"📘 Vocab Size":255000,"Tokenizer Class":"CohereTokenizerFast"}
|
7 |
+
{"👳 Tokenize Tashkeel":"❌","📛 Models":"CohereForAI\/c4ai-command-r-plus","🪺 Fertility Score":2.154,"➕ Total Number of Tokens":1658463,"📘 Vocab Size":255000,"Tokenizer Class":"CohereTokenizerFast"}
|
8 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"unsloth\/gemma-2b-bnb-4bit","🪺 Fertility Score":2.199,"➕ Total Number of Tokens":1692826,"📘 Vocab Size":256000,"Tokenizer Class":"GemmaTokenizerFast"}
|
9 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"NousResearch\/Meta-Llama-3-8B","🪺 Fertility Score":2.374,"➕ Total Number of Tokens":1827816,"📘 Vocab Size":128000,"Tokenizer Class":"PreTrainedTokenizerFast"}
|
10 |
+
{"👳 Tokenize Tashkeel":"❌","📛 Models":"Qwen\/Qwen1.5-7B-Chat","🪺 Fertility Score":2.444,"➕ Total Number of Tokens":1881958,"📘 Vocab Size":151643,"Tokenizer Class":"Qwen2TokenizerFast"}
|
11 |
+
{"👳 Tokenize Tashkeel":"❌","📛 Models":"Qwen\/Qwen1.5-110B-Chat","🪺 Fertility Score":2.444,"➕ Total Number of Tokens":1881958,"📘 Vocab Size":151643,"Tokenizer Class":"Qwen2TokenizerFast"}
|
12 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"FreedomIntelligence\/AceGPT-13B","🪺 Fertility Score":5.46,"➕ Total Number of Tokens":4203685,"📘 Vocab Size":32000,"Tokenizer Class":"LlamaTokenizerFast"}
|
13 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"microsoft\/Phi-3-mini-128k-instruct","🪺 Fertility Score":5.46,"➕ Total Number of Tokens":4203685,"📘 Vocab Size":32000,"Tokenizer Class":"LlamaTokenizerFast"}
|
14 |
+
{"👳 Tokenize Tashkeel":"✅","📛 Models":"01-ai\/Yi-1.5-34B-Chat","🪺 Fertility Score":6.674,"➕ Total Number of Tokens":5138447,"📘 Vocab Size":64000,"Tokenizer Class":"LlamaTokenizerFast"}
|