Commit
·
b4ccea2
1
Parent(s):
db310d7
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: apache-2.0
|
| 4 |
+
datasets:
|
| 5 |
+
- Technotech/sd-prompt-instruct
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
+
tags:
|
| 9 |
+
- stable-diffusion
|
| 10 |
+
- instruct
|
| 11 |
+
- magic-prompt
|
| 12 |
+
- natural language inference
|
| 13 |
+
---
|
| 14 |
+
# Stable Diffusion Prompt Instruct 3B GGML (OpenLlama v2 3B)
|
| 15 |
+
|
| 16 |
+
Trained for 0.4 epochs (test) on [Technotech/sd-prompt-instruct](https://huggingface.co/datasets/Technotech/sd-prompt-instruct).
|
| 17 |
+
|
| 18 |
+
## Prompt Format
|
| 19 |
+
|
| 20 |
+
```
|
| 21 |
+
### Instruction: {prompt}
|
| 22 |
+
### Response: {response}
|
| 23 |
+
```
|
| 24 |
+
|
| 25 |
+
## Formats
|
| 26 |
+
|
| 27 |
+
At the moment, k-quants are not compatible with OpenLlama v2 3B, which this model is fine tuned from.
|
| 28 |
+
|
| 29 |
+
| Quant | Name | Size |
|
| 30 |
+
| ----- | ----- | ----- |
|
| 31 |
+
| `q4_0` | `sd-prompt-instruct-ggml.q4_0.bin` | `(1.93 GB)`
|
| 32 |
+
| `q4_1` | `sd-prompt-instruct-ggml.q4_1.bin` | `(2.14 GB)`
|
| 33 |
+
| `q5_0` | `sd-prompt-instruct-ggml.q5_0.bin` | `(2.36 GB)`
|
| 34 |
+
| `q5_1` | `sd-prompt-instruct-ggml.q5_1.bin` | `(2.57 GB)`
|