-
```
Backup pages storage takes a lot of memory. We should compress them.
The compression should be lossless of course. Compression should be fast
because it's used very often when you draw. Decompres…
-
llama3-sft.yaml
```yaml
### model
model_name_or_path: models/llama-3-8b-Instruct
### method
stage: sft
do_train: true
finetuning_type: full
deepspeed: examples/deepspeed/ds_z3_config.json
…
-
Add clearance between texture rectangles when bin packing. This will allow for easy reduction when it is possible.
-
For some mod projects with several file variants, it can be inconvenient to store duplicate files across multiple projects.
I would love the option to select files to be packed, which would allow q…
-
### Is this a new feature, an improvement, or a change to existing functionality?
Improvement
### How would you describe the priority of this feature request
Medium
### Please provide a clear desc…
-
### Is your feature request related to a problem? Please describe.
There is severely the lack of nodes that can calculate a high precision type such as a decimal type, and would be nice to have the…
-
Not using msgpack, but using the new binary packing/unpacking modules to make the savedata code more robust and less complex.
-
### Please check that this issue hasn't been reported before.
- [X] I searched previous [Bug Reports](https://github.com/OpenAccess-AI-Collective/axolotl/labels/bug) didn't find any similar reports.
…
-
Hi there,
Could you please have a look at two packs below. The first one is working with keepFlat flag enabled and another one without. Both variants use the same list of boxes (please see jsons).
…
pikil updated
3 years ago
-
```
Backup pages storage takes a lot of memory. We should compress them.
The compression should be lossless of course. Compression should be fast
because it's used very often when you draw. Decompres…