Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,11 @@
|
|
| 1 |
---
|
| 2 |
license: llama2
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: llama2
|
| 3 |
---
|
| 4 |
+
|
| 5 |
+
This is an interleaved merge of [Xwin-longLORA-70b-rope8-32k-fp16](https://huggingface.co/grimulkan/Xwin-longLORA-70b-rope8-32k-fp16) and [Euryale-1.3-longLORA-70b-rope8-32k-fp16](https://huggingface.co/grimulkan/Euryale-1.3-longLORA-70b-rope8-32k-fp16), replacing the embed and norm layers as described in the [LongLoRA repo](https://github.com/dvlab-research/LongLoRA), using the same merge formula as alpindale's [goliath-120b](https://huggingface.co/alpindale/goliath-120b).
|
| 6 |
+
|
| 7 |
+
There is no additional fine-tuning. The resulting model seems to not be broken... you can test whether it is truly the original model + 32K capability (use linear rope scaling 8).
|
| 8 |
+
|
| 9 |
+
A 6-bit EXL2 quantization is available [here](https://huggingface.co/grimulkan/Goliath-longLORA-120b-rope8-2k-6bpw_h8_exl2).
|
| 10 |
+
|
| 11 |
+
See [this discussion](https://huggingface.co/grimulkan/aurelian-v0.5-70b-rope8-32K-fp16/discussions/2) for how the original 70B merges were created with longLORA.
|