Lora for AlphaMonarch-7B which allows you to convert a given model sentence into a set of danbooru tags format: Alpaca train format: Alpaca
This is the first time I have had this lora done and I am quite happy with the result. This lora was trained on full model AlphaMonarch-7B. with rank of 8 for 3 epochs There were 15k instructions in the dataset that had examples of how to convert from sentences to tags.
I use following prompt
Here are examples following these rules:
Input: A boy with glasses is drawing in his sketchbook at a park during autumn.
Output: 1boy, solo, glasses, drawing, sketchbook, park, autumn
Input: Two girls are jogging together around a lake at sunrise; one has red hair and the other has brown.
Output: 2girls, jogging together, lake, sunrise, red hair girl, brown hair girl
Input: A girl with cat ears, wearing a hoodie and with long white hair and red eyes walks arm in arm with a girl in a black bikini with brown hair and green eyes
Output: 2girls, walking, holding hands\n1girl, cat ears, hoodie, long hair, white hair, red eyes\n1girl, black bikini, brown hair, green eyes
Input: 3 girls walking in the park, one in a hoodie with green hair, one with red hair and a dress, 3 in shorts and a top with white hair
Output: 3girls, walking in park\n1girl, green hair, hoodie\n1girl, red hair, dress\n1girl, white hair, shorts, top
Important note:
- Always start with the number of characters and their gender (e.g., "1girl," "3boys").
- If a character is alone, add the "solo" tag.
- Describe physical attributes such as hair color or clothing (e.g., "pink hair," "gothic lolita fashion").
- Describe actions of characters if any (e.g., “playing violin,” “studying”).
- Mention setting details like location and background objects (e.g.,”music room”, “classical instruments”).
- Keep each tag separated by a comma.
- All words should be tags that can be tagged to danboororu, they should be as informative as possible and represent a single entity or action.
- From the verbs, select words that can denote place
- If there are multiple people in the query, separate them with a \n. The first line is the number of people and location. And each following line is a description of one person.
- Answer only on English