Commit
·
4717edd
1
Parent(s):
03b190f
Update README.md
Browse files
README.md
CHANGED
|
@@ -10,14 +10,6 @@ Introducing EverythingLM, a llama-2 based, general-purpose 13b model with 16k co
|
|
| 10 |
|
| 11 |
The model is completely uncensored.
|
| 12 |
|
| 13 |
-
### GGML quants:
|
| 14 |
-
soon
|
| 15 |
-
|
| 16 |
-
Make sure to use correct rope scaling settings:
|
| 17 |
-
`-c 16384 --rope-freq-base 10000 --rope-freq-scale 0.25`
|
| 18 |
-
### GPTQ quants:
|
| 19 |
-
soon
|
| 20 |
-
|
| 21 |
### Notable features:
|
| 22 |
- Automatically triggered CoT reasoning.
|
| 23 |
- Verbose and detailed replies.
|
|
@@ -38,4 +30,4 @@ Training took about 2.5 hours using QLoRa on 1xA100, so this model can be recrea
|
|
| 38 |
### Future plans:
|
| 39 |
- Native finetune.
|
| 40 |
- Other model sizes.
|
| 41 |
-
- Test some model merges using this model.
|
|
|
|
| 10 |
|
| 11 |
The model is completely uncensored.
|
| 12 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
### Notable features:
|
| 14 |
- Automatically triggered CoT reasoning.
|
| 15 |
- Verbose and detailed replies.
|
|
|
|
| 30 |
### Future plans:
|
| 31 |
- Native finetune.
|
| 32 |
- Other model sizes.
|
| 33 |
+
- Test some model merges using this model.
|