sam-mosaic
commited on
Commit
•
f3359cd
1
Parent(s):
bd1748e
add formatting example
Browse files
README.md
CHANGED
@@ -98,6 +98,31 @@ from transformers import AutoTokenizer
|
|
98 |
tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
|
99 |
```
|
100 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
101 |
## Model Description
|
102 |
|
103 |
The architecture is a modification of a standard decoder-only transformer.
|
|
|
98 |
tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
|
99 |
```
|
100 |
|
101 |
+
### Formatting
|
102 |
+
|
103 |
+
This model was trained on data formatted in the dolly-15k format:
|
104 |
+
|
105 |
+
```python
|
106 |
+
INSTRUCTION_KEY = "### Instruction:"
|
107 |
+
RESPONSE_KEY = "### Response:"
|
108 |
+
INTRO_BLURB = "Below is an instruction that describes a task. Write a response that appropriately completes the request."
|
109 |
+
PROMPT_FOR_GENERATION_FORMAT = """{intro}
|
110 |
+
{instruction_key}
|
111 |
+
{instruction}
|
112 |
+
{response_key}
|
113 |
+
""".format(
|
114 |
+
intro=INTRO_BLURB,
|
115 |
+
instruction_key=INSTRUCTION_KEY,
|
116 |
+
instruction="{instruction}",
|
117 |
+
response_key=RESPONSE_KEY,
|
118 |
+
)
|
119 |
+
|
120 |
+
example = "James decides to run 3 sprints 3 times a week. He runs 60 meters each sprint. How many total meters does he run a week? Explain before answering."
|
121 |
+
fmt_ex = PROMPT_FOR_GENERATION_FORMAT.format(instruction=example)
|
122 |
+
```
|
123 |
+
|
124 |
+
In the above example, `fmt_ex` is ready to be tokenized and sent through the model.
|
125 |
+
|
126 |
## Model Description
|
127 |
|
128 |
The architecture is a modification of a standard decoder-only transformer.
|