Datasets:
andreaskoepf
commited on
Commit
•
56e1636
1
Parent(s):
670f5bc
Update README.md
Browse files
README.md
CHANGED
@@ -111,6 +111,7 @@ tags:
|
|
111 |
- human-feedback
|
112 |
size_categories:
|
113 |
- 10K<n<100K
|
|
|
114 |
---
|
115 |
|
116 |
# OpenAssistant Conversations Dataset (OASST1)
|
@@ -129,6 +130,8 @@ corpus consisting of 161,443 messages distributed across 66,497 conversation tre
|
|
129 |
35 different languages, annotated with 461,292 quality ratings. The corpus is a product
|
130 |
of a worldwide crowd-sourcing effort involving over 13,500 volunteers.
|
131 |
|
|
|
|
|
132 |
### Dataset Structure
|
133 |
|
134 |
This dataset contains demonstrations of human-assistant conversations which were collected
|
@@ -144,46 +147,51 @@ conversation threads from prompt to leaf node in a conversation tree are stricly
|
|
144 |
between "assistant" and "prompter".
|
145 |
|
146 |
Please refer to [oasst-data](https://github.com/LAION-AI/Open-Assistant/tree/main/oasst-data) for
|
147 |
-
details about the data structure and python code to
|
148 |
-
jsonl files.
|
149 |
|
150 |
|
151 |
## Main Dataset Files
|
152 |
|
153 |
-
Data is provided either as nested messages in conversation trees
|
154 |
-
|
155 |
-
The type of file can be inferred from the file name extension:
|
156 |
-
- `.trees.jsonl.gz`: Conversation trees with nested messages
|
157 |
-
- `.messages.jsonl.gz`: Flat list of messages
|
158 |
|
159 |
-
Full conversation trees can be reconstructed from
|
160 |
-
and `message_id` properties to identify
|
161 |
-
|
162 |
-
|
163 |
|
164 |
### Ready For Export Trees
|
165 |
|
166 |
-
This set of exported trees contains all trees in `ready_for_export` state without spam
|
167 |
-
and deleted messages. For supervised fine-tuning & reward model training this file is ideal.
|
168 |
-
|
169 |
```
|
170 |
2023-04-12_oasst_ready.trees.jsonl.gz 10364 trees with 88838 total messages
|
171 |
2023-04-12_oasst_ready.messages.jsonl.gz 88838 messages
|
172 |
```
|
|
|
|
|
173 |
|
174 |
-
### All Trees
|
175 |
|
|
|
176 |
|
177 |
```
|
178 |
2023-04-12_oasst_all.trees.jsonl.gz 66497 trees with 161443 total messages
|
179 |
2023-04-12_oasst_all.messages.jsonl.gz 161443 messages
|
180 |
```
|
|
|
|
|
181 |
|
182 |
### Supplemental Exports: Spam & Prompts
|
183 |
|
184 |
```
|
|
|
|
|
|
|
|
|
185 |
|
|
|
|
|
|
|
186 |
```
|
|
|
187 |
|
188 |
|
189 |
### Languages
|
@@ -231,4 +239,4 @@ OpenAssistant Conversations incorporates 35 different languages with a distribut
|
|
231 |
<li>Esperanto: 59</li>
|
232 |
<li>Slovak: 19</li>
|
233 |
</ul>
|
234 |
-
</details>
|
|
|
111 |
- human-feedback
|
112 |
size_categories:
|
113 |
- 10K<n<100K
|
114 |
+
pretty_name: OpenAssistant Conversations
|
115 |
---
|
116 |
|
117 |
# OpenAssistant Conversations Dataset (OASST1)
|
|
|
130 |
35 different languages, annotated with 461,292 quality ratings. The corpus is a product
|
131 |
of a worldwide crowd-sourcing effort involving over 13,500 volunteers.
|
132 |
|
133 |
+
Please our paper for further details.
|
134 |
+
|
135 |
### Dataset Structure
|
136 |
|
137 |
This dataset contains demonstrations of human-assistant conversations which were collected
|
|
|
147 |
between "assistant" and "prompter".
|
148 |
|
149 |
Please refer to [oasst-data](https://github.com/LAION-AI/Open-Assistant/tree/main/oasst-data) for
|
150 |
+
details about the data structure and python code to read and write jsonl files containing oasst objects.
|
|
|
151 |
|
152 |
|
153 |
## Main Dataset Files
|
154 |
|
155 |
+
Data is provided either as nested messages in conversation trees (extension `.trees.jsonl.gz`)
|
156 |
+
or as flat list of messages (extension `.messages.jsonl.gz`).
|
|
|
|
|
|
|
157 |
|
158 |
+
Full conversation trees can be reconstructed from flat messages using the `parent_id`
|
159 |
+
and `message_id` properties to identify their parent-child relationship. The `message_tree_id`
|
160 |
+
and `tree_state` properties (only present in flat messages files) can be used to find all
|
161 |
+
all messages of a message tree or to select trees by their state.
|
162 |
|
163 |
### Ready For Export Trees
|
164 |
|
|
|
|
|
|
|
165 |
```
|
166 |
2023-04-12_oasst_ready.trees.jsonl.gz 10364 trees with 88838 total messages
|
167 |
2023-04-12_oasst_ready.messages.jsonl.gz 88838 messages
|
168 |
```
|
169 |
+
Trees in `ready_for_export` state without spam and deleted messages including message labels.
|
170 |
+
The oasst_ready-trees file is normally sufficient for supervised fine-tuning (SFT) & reward model (RM) training.
|
171 |
|
|
|
172 |
|
173 |
+
### All Trees
|
174 |
|
175 |
```
|
176 |
2023-04-12_oasst_all.trees.jsonl.gz 66497 trees with 161443 total messages
|
177 |
2023-04-12_oasst_all.messages.jsonl.gz 161443 messages
|
178 |
```
|
179 |
+
All trees including those in states `prompt_lottery_waiting`, `aborted_low_grade`, `halted_by_moderator`.
|
180 |
+
|
181 |
|
182 |
### Supplemental Exports: Spam & Prompts
|
183 |
|
184 |
```
|
185 |
+
2023-04-12_oasst_spam.messages.jsonl.gz
|
186 |
+
```
|
187 |
+
Messages which were deleted or have a negative review result (`"review_result": false`).
|
188 |
+
Beside low quality a frequent reason for message deletion is a wrong language tag.
|
189 |
|
190 |
+
|
191 |
+
```
|
192 |
+
2023-04-12_oasst_prompts.messages.jsonl.gz
|
193 |
```
|
194 |
+
All non-deleted initial prompt messages with positile spam review result of trees in `ready_for_export` or `prompt_lottery_waiting` state.
|
195 |
|
196 |
|
197 |
### Languages
|
|
|
239 |
<li>Esperanto: 59</li>
|
240 |
<li>Slovak: 19</li>
|
241 |
</ul>
|
242 |
+
</details>
|