Update README.md

README.md

@@ -14,16 +14,25 @@ inference: false
 ### This repository contains quantized conversions of the current Pygmalion 6B checkpoints.
 *For use with frontends that support GGML quantized GPT-J models, such as KoboldCpp and Oobabooga (with the CTransformers loader).*
 
+*Last updated on 2023-09-26.*
+
+**Description:**
+- The `pygmalion-6b-main` files are quantized from the main branch of Pygmalion 6B from 2023-01-13. Also known as "experiment 2".
+- The `pygmalion-6b-dev` files are quantized from the dev branch of Pygmalion 6B from 2023-03-12. Also known as "part 4/10 of experiment 7".
+
+**RAM usage:**
 Model | Startup RAM usage (KoboldCpp) | Startup RAM usage (Oobabooga)
 :--:|:--:|:--:
-pygmalion-6b-main.q4_0.bin | Placeholder | Placeholder
-pygmalion-6b-main.q4_1.bin | Placeholder | Placeholder
-pygmalion-6b-main.q5_0.bin | Placeholder | Placeholder
-pygmalion-6b-main.q5_1.bin | Placeholder | Placeholder
-pygmalion-6b-main.q8_0.bin | Placeholder | Placeholder
-pygmalion-6b-main.f16.bin | Placeholder | Placeholder
-
-The original model card can be found below.
+pygmalion-6b-main.q4_0.bin | 3.7 GiB | 3.7 GiB
+pygmalion-6b-main.q4_1.bin | 4.1 GiB | 4.1 GiB
+pygmalion-6b-main.q5_0.bin | 4.4 GiB | 4.4 GiB
+pygmalion-6b-main.q5_1.bin | 4.8 GiB | 4.8 GiB
+pygmalion-6b-main.q8_0.bin | 6.5 GiB | 6.6 GiB
+
+**Notes:**
+- ggerganov/ggml [[8ca2c19]](https://github.com/ggerganov/ggml/tree/8ca2c19a3bb8622954d858fbf6383522684eaf34)'s gpt-j conversion script was used for conversion and quantization. First they were converted to f16 ggml files, then quantized.
+
+The original model can be found [here](https://huggingface.co/PygmalionAI/pygmalion-6b), and the original model card can be found below.
 
 * * *