Update README.md
Browse files
README.md
CHANGED
|
@@ -11,12 +11,13 @@ inference: false
|
|
| 11 |
|
| 12 |
*(Not to be confused with [Pygmalion 13B](https://huggingface.co/TehVenom/Pygmalion-13b-GGML).)*
|
| 13 |
|
| 14 |
-
# Converted with ggerganov/ggml's gpt-neox conversion script, and tested with KoboldCpp.
|
| 15 |
-
## *(I can't promise that this will work with other frontends, if at all; I haven't had the most success myself. Use at your own risk!)*
|
| 16 |
-
|
| 17 |
This is converted and quantized from [Pygmalion 1.3B](https://huggingface.co/PygmalionAI/pygmalion-1.3b), based on [an earlier version](https://huggingface.co/EleutherAI/pythia-1.4b-deduped-v0) of Pythia 1.4B Deduped.
|
| 18 |
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
Model | Initial RAM
|
| 21 |
:--:|:--:
|
| 22 |
ggml-pygmalion-1.3b-q4_0.bin | 1.1 GiB
|
|
|
|
| 11 |
|
| 12 |
*(Not to be confused with [Pygmalion 13B](https://huggingface.co/TehVenom/Pygmalion-13b-GGML).)*
|
| 13 |
|
|
|
|
|
|
|
|
|
|
| 14 |
This is converted and quantized from [Pygmalion 1.3B](https://huggingface.co/PygmalionAI/pygmalion-1.3b), based on [an earlier version](https://huggingface.co/EleutherAI/pythia-1.4b-deduped-v0) of Pythia 1.4B Deduped.
|
| 15 |
|
| 16 |
+
Notes:
|
| 17 |
+
- Converted with ggerganov/ggml's gpt-neox conversion script, and tested with KoboldCpp.
|
| 18 |
+
- I can't promise that this will work with other frontends, if at all. I've had problems with the tokenizer. Could be related to the ggml implementation of GPT-NeoX [(source)](https://github.com/ggerganov/ggml/tree/master/examples/gpt-neox#notes).
|
| 19 |
+
|
| 20 |
+
### RAM USAGE (on KoboldCpp w/ OpenBLAS)
|
| 21 |
Model | Initial RAM
|
| 22 |
:--:|:--:
|
| 23 |
ggml-pygmalion-1.3b-q4_0.bin | 1.1 GiB
|