ngxson
|
42415a4874
|
auto scale
|
2024-07-15 11:41:18 +02:00 |
Francis Couture-Harpin
|
8956543c09
|
convert_hf : simplify modify_tensors for InternLM2
* convert_lora : lazy conversion
* llama : load and use alpha from LoRA adapters
|
2024-07-15 02:48:24 -04:00 |
|
Francis Couture-Harpin
|
9d96328bdf
|
convert_lora : MoE LoRA conversion support
* convert_lora : prefer safetensors, similarly to convert_hf
|
2024-07-11 21:32:38 -04:00 |
|
ngxson
|
713665db2e
|
fix types
|
2024-07-10 00:36:52 +02:00 |
|
ngxson
|
ee2b35c65f
|
conversion: only allow selected models
|
2024-07-10 00:23:07 +02:00 |
|
ngxson
|
95b3eb057b
|
fix outfile
|
2024-07-08 22:05:35 +02:00 |
|
ngxson
|
7a83f200d3
|
fix ftype
|
2024-07-08 21:55:41 +02:00 |
|
ngxson
|
84288ff9f7
|
add f16 convert
|
2024-07-08 17:05:17 +02:00 |
|
ngxson
|
847135aaa2
|
add convert script
|
2024-07-08 16:35:27 +02:00 |