Commit
•
da042fb
1
Parent(s):
3fa17ab
Update README.md
Browse files
README.md
CHANGED
@@ -8,6 +8,32 @@ language:
|
|
8 |
tags:
|
9 |
- completion
|
10 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
## Training procedure
|
12 |
|
13 |
|
|
|
8 |
tags:
|
9 |
- completion
|
10 |
---
|
11 |
+
## Info
|
12 |
+
|
13 |
+
Magic prompt completion model trained on a dataset 70k Stable Diffusion prompts. Base model: TinyStories-33M.
|
14 |
+
|
15 |
+
Model seems to be pretty decent for 33M params, but it clearly lacks much of an understanding of pretty much anything. Still, considering the size, I think it's decent. Whether you would use this over a small GPT-2 based model is up to you.
|
16 |
+
|
17 |
+
## Examples
|
18 |
+
|
19 |
+
Generation settings: `max_new_tokens=40, do_sample=True, temperature=2.0, num_beams=10, repetition_penalty=1.2, top_k=40, top_p=0.75, eos_token_id=tokenizer.eos_token_id` (there may be better settings).
|
20 |
+
|
21 |
+
(Bold text is generated by the model)
|
22 |
+
|
23 |
+
"A close shot of a bird in a jungle, **with two legs, with long hair on a tall, long brown body, long white skin, sharp teeth, high bones, digital painting, artstation, concept art, illustration by wlop,**"
|
24 |
+
|
25 |
+
"Camera shot of **a strange young girl wearing a cloak, wearing a mask in clothes, with long curly hair, long hair, black eyes, dark skin, white teeth, long brown eyes eyes, big eyes, sharp**"
|
26 |
+
|
27 |
+
"An illustration of a house, stormy weather, **sun, moonlight, night, concept art, 4 k, wlop, by wlop, by jose stanley, ilya kuvshinov, sprig**"
|
28 |
+
|
29 |
+
"A field of flowers, camera shot, 70mm lens, **fantasy, intricate, highly detailed, artstation, concept art, sharp focus, illustration, illustration, artgerm jake daggaws, artgerm and jaggodieie brad**"
|
30 |
+
|
31 |
+
## Training config
|
32 |
+
|
33 |
+
- Rank 16 LoRA
|
34 |
+
- Trained on Gustavosta/Stable-Diffusion-Prompts for 10 epochs
|
35 |
+
- Batch size of 64
|
36 |
+
|
37 |
## Training procedure
|
38 |
|
39 |
|