Update README.md
Browse files
README.md
CHANGED
@@ -14,6 +14,197 @@ This tokenizer was trained on 2M samples from:
|
|
14 |
|
15 |
## How to use
|
16 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
```python
|
18 |
from transformers import AutoTokenizer
|
19 |
|
@@ -128,4 +319,5 @@ What's wrong with Cheems? What treatment does he need?<|end_of_text|>
|
|
128 |
|
129 |
<|start_header_id|>assistant<|end_header_id|>
|
130 |
Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters.<|end_of_text|>
|
131 |
-
```
|
|
|
|
14 |
|
15 |
## How to use
|
16 |
|
17 |
+
<details>
|
18 |
+
<summary>Only conversation</summary>
|
19 |
+
|
20 |
+
```python
|
21 |
+
from transformers import AutoTokenizer
|
22 |
+
|
23 |
+
tokenizer = AutoTokenizer.from_pretrained("SmallDoge/Doge-tokenizer")
|
24 |
+
|
25 |
+
conversation = [
|
26 |
+
{"role": "system", "content": "You are a helpful assistant."},
|
27 |
+
{"role": "user", "content": "What's wrong with Cheems? What treatment does he need?"},
|
28 |
+
{"role": "assistant", "content": "Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters."},
|
29 |
+
]
|
30 |
+
|
31 |
+
inputs = tokenizer.apply_chat_template(
|
32 |
+
conversation=conversation,
|
33 |
+
tokenize=False,
|
34 |
+
)
|
35 |
+
print(inputs)
|
36 |
+
```
|
37 |
+
```shell
|
38 |
+
<|begin_of_text|><|start_header_id|>system<|end_header_id|>
|
39 |
+
Cutting Knowledge Date: December 2024
|
40 |
+
Today Date: December 2025
|
41 |
+
You are a helpful assistant.<|end_of_text|>
|
42 |
+
|
43 |
+
<|start_header_id|>user<|end_header_id|>
|
44 |
+
What's wrong with Cheems? What treatment does he need?<|end_of_text|>
|
45 |
+
|
46 |
+
<|start_header_id|>assistant<|end_header_id|>
|
47 |
+
Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters.<|end_of_text|>
|
48 |
+
```
|
49 |
+
</details>
|
50 |
+
|
51 |
+
<details>
|
52 |
+
<summary>With documents</summary>
|
53 |
+
|
54 |
+
```python
|
55 |
+
from transformers import AutoTokenizer
|
56 |
+
|
57 |
+
tokenizer = AutoTokenizer.from_pretrained("SmallDoge/Doge-tokenizer")
|
58 |
+
|
59 |
+
documents = [
|
60 |
+
{
|
61 |
+
"title": "Cheems's case",
|
62 |
+
"text": "Cheems is a 233-year-old alchemist, but he is very kidney deficient."
|
63 |
+
},
|
64 |
+
]
|
65 |
+
|
66 |
+
conversation = [
|
67 |
+
{"role": "system", "content": "You are a helpful assistant."},
|
68 |
+
{"role": "user", "content": "What's wrong with Cheems? What treatment does he need?"},
|
69 |
+
{"role": "assistant", "content": "Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters."},
|
70 |
+
]
|
71 |
+
|
72 |
+
inputs = tokenizer.apply_chat_template(
|
73 |
+
documents=documents,
|
74 |
+
conversation=conversation,
|
75 |
+
tokenize=False,
|
76 |
+
)
|
77 |
+
print(inputs)
|
78 |
+
```
|
79 |
+
```shell
|
80 |
+
<|begin_of_text|><|start_header_id|>system<|end_header_id|>
|
81 |
+
Cutting Knowledge Date: December 2024
|
82 |
+
Today Date: December 2025
|
83 |
+
You are a helpful assistant.<|end_of_text|>
|
84 |
+
|
85 |
+
<|start_header_id|>user<|end_header_id|>
|
86 |
+
Given the following documents, please use them to answer the user's question.
|
87 |
+
|
88 |
+
Title: Cheems' case
|
89 |
+
Content: Cheems is a 233-year-old alchemist, but he is very kidney deficient.
|
90 |
+
|
91 |
+
If the documents don't contain relevant information, rely on your general knowledge but acknowledge when you're doing so.
|
92 |
+
|
93 |
+
What's wrong with Cheems? What treatment does he need?<|end_of_text|>
|
94 |
+
|
95 |
+
<|start_header_id|>assistant<|end_header_id|>
|
96 |
+
Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters.<|end_of_text|>
|
97 |
+
```
|
98 |
+
</details>
|
99 |
+
|
100 |
+
<details>
|
101 |
+
<summary>With tools</summary>
|
102 |
+
|
103 |
+
```python
|
104 |
+
from transformers import AutoTokenizer
|
105 |
+
|
106 |
+
tokenizer = AutoTokenizer.from_pretrained("SmallDoge/Doge-tokenizer")
|
107 |
+
|
108 |
+
tools = [
|
109 |
+
{
|
110 |
+
"name": "recommend_treatment",
|
111 |
+
"description": "Treatment is recommended based on patient symptoms and diagnosis.",
|
112 |
+
"parameters": {
|
113 |
+
"type": "object",
|
114 |
+
"properties": {
|
115 |
+
"diagnosis": {
|
116 |
+
"type": "string",
|
117 |
+
"description": "Patient diagnostic results"
|
118 |
+
},
|
119 |
+
"severity": {
|
120 |
+
"type": "string",
|
121 |
+
"enum": ["mild", "moderate", "severe"],
|
122 |
+
"description": "Severity of the condition, can be mild, moderate, or severe."
|
123 |
+
}
|
124 |
+
},
|
125 |
+
"required": ["diagnosis", "severity"]
|
126 |
+
}
|
127 |
+
}
|
128 |
+
]
|
129 |
+
|
130 |
+
conversation = [
|
131 |
+
{"role": "system", "content": "You are a helpful assistant."},
|
132 |
+
{"role": "user", "content": "What's wrong with Cheems? What treatment does he need?"},
|
133 |
+
{"role": "assistant", "tool_calls": [
|
134 |
+
{
|
135 |
+
"id": "call_1",
|
136 |
+
"function": {
|
137 |
+
"name": "recommend_treatment",
|
138 |
+
"arguments": "{\"diagnosis\": \"Deficiency of kidney\", \"severity\": \"severe\"}"
|
139 |
+
}
|
140 |
+
}
|
141 |
+
]},
|
142 |
+
{"role": "tool", "tool_call_id": "call_1", "content": "For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters."},
|
143 |
+
{"role": "assistant", "content": "Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters."},
|
144 |
+
]
|
145 |
+
|
146 |
+
inputs = tokenizer.apply_chat_template(
|
147 |
+
tools=tools,
|
148 |
+
conversation=conversation,
|
149 |
+
tokenize=False,
|
150 |
+
)
|
151 |
+
print(inputs)
|
152 |
+
```
|
153 |
+
```shell
|
154 |
+
<|begin_of_text|><|start_header_id|>system<|end_header_id|>
|
155 |
+
Environment: ipython
|
156 |
+
Cutting Knowledge Date: December 2024
|
157 |
+
Today Date: December 2025
|
158 |
+
You are a helpful assistant.<|end_of_text|>
|
159 |
+
|
160 |
+
<|start_header_id|>user<|end_header_id|>
|
161 |
+
Given the following functions, please respond with a JSON for a function call with its proper arguments that best answers the given prompt.
|
162 |
+
|
163 |
+
{
|
164 |
+
"name": "recommend_treatment",
|
165 |
+
"description": "Treatment is recommended based on patient symptoms and diagnosis.",
|
166 |
+
"parameters": {
|
167 |
+
"type": "object",
|
168 |
+
"properties": {
|
169 |
+
"diagnosis": {
|
170 |
+
"type": "string",
|
171 |
+
"description": "Patient diagnostic results"
|
172 |
+
},
|
173 |
+
"severity": {
|
174 |
+
"type": "string",
|
175 |
+
"enum": [
|
176 |
+
"mild",
|
177 |
+
"moderate",
|
178 |
+
"severe"
|
179 |
+
],
|
180 |
+
"description": "Severity of the condition, can be mild, moderate, or severe."
|
181 |
+
}
|
182 |
+
},
|
183 |
+
"required": [
|
184 |
+
"diagnosis",
|
185 |
+
"severity"
|
186 |
+
]
|
187 |
+
}
|
188 |
+
}
|
189 |
+
|
190 |
+
Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}. Do not use variables.
|
191 |
+
|
192 |
+
What's wrong with Cheems? What treatment does he need?<|end_of_text|>
|
193 |
+
|
194 |
+
<|start_header_id|>assistant<|end_header_id|>
|
195 |
+
{"name": "recommend_treatment", "parameters": "{\"diagnosis\": \"Deficiency of kidney\", \"severity\": \"severe\"}"}<|end_of_text|>
|
196 |
+
|
197 |
+
<|start_header_id|>ipython<|end_header_id|>
|
198 |
+
"For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters."<|end_of_text|>
|
199 |
+
|
200 |
+
<|start_header_id|>assistant<|end_header_id|>
|
201 |
+
Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters.<|end_of_text|>
|
202 |
+
```
|
203 |
+
</details>
|
204 |
+
|
205 |
+
<details>
|
206 |
+
<summary>With documents and tools</summary>
|
207 |
+
|
208 |
```python
|
209 |
from transformers import AutoTokenizer
|
210 |
|
|
|
319 |
|
320 |
<|start_header_id|>assistant<|end_header_id|>
|
321 |
Based on Cheems' symptoms, I recommend the following treatment: For Cheems with severe kidney deficiency, it is recommended to eat more leeks and oysters.<|end_of_text|>
|
322 |
+
```
|
323 |
+
</details>
|