Hugging Face's logo

pszemraj
/
distilgpt2-magicprompt-SD

Text Generation
Transformers
PyTorch
Safetensors
gpt2
Generated from Trainer
stable diffusion
diffusion
text2image
prompt augment
prompt engineering
text-generation-inference
Inference Endpoints
Model card Files Community
2
pszemraj commited on
Commit
206f6de
•
1 Parent(s): 7b9d6f3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -1
README.md CHANGED
@@ -27,13 +27,21 @@ parameters:
27
 
28
  Generate/augment your prompt, stable diffusion style.
29
 
30
-
31
  This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on the Gustavosta/Stable-Diffusion-Prompts dataset.
32
  It achieves the following results on the evaluation set:
33
  - Loss: 1.3089
34
  - eval_steps_per_second = 17.201
35
  - perplexity = 3.7022
36
 
 
 
 
 
 
 
 
 
 
37
 
38
  ## Training and evaluation data
39
 
 
27
 
28
  Generate/augment your prompt, stable diffusion style.
29
 
 
30
  This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on the Gustavosta/Stable-Diffusion-Prompts dataset.
31
  It achieves the following results on the evaluation set:
32
  - Loss: 1.3089
33
  - eval_steps_per_second = 17.201
34
  - perplexity = 3.7022
35
 
36
+ ## example
37
+
38
+ ![prompt](https://i.imgur.com/DKHoLtB.png)
39
+
40
+ Results in (_DALL-E, but you get the idea_):
41
+
42
+ ![sentient cheese astronaut](https://i.ibb.co/zrCf4BK/DALL-E-2022-11-09-12-17-40-sentient-cheese-robot-astronaut-highly-detailed-concept-art-eric-zener-el.png)
43
+
44
+ this `distilgpt2` version is probably small/fast enough to be used locally on CPU!
45
 
46
  ## Training and evaluation data
47