Isabelle Cudlitz
commited on
Commit
·
f5df9d4
1
Parent(s):
84304e9
Upload model
Browse files- README.md +18 -0
- adapter_model.bin +1 -1
README.md
CHANGED
|
@@ -1044,4 +1044,22 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 1044 |
### Framework versions
|
| 1045 |
|
| 1046 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1047 |
- PEFT 0.6.2
|
|
|
|
| 1044 |
### Framework versions
|
| 1045 |
|
| 1046 |
|
| 1047 |
+
- PEFT 0.6.2
|
| 1048 |
+
## Training procedure
|
| 1049 |
+
|
| 1050 |
+
|
| 1051 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 1052 |
+
- load_in_8bit: False
|
| 1053 |
+
- load_in_4bit: True
|
| 1054 |
+
- llm_int8_threshold: 6.0
|
| 1055 |
+
- llm_int8_skip_modules: None
|
| 1056 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 1057 |
+
- llm_int8_has_fp16_weight: False
|
| 1058 |
+
- bnb_4bit_quant_type: fp4
|
| 1059 |
+
- bnb_4bit_use_double_quant: True
|
| 1060 |
+
- bnb_4bit_compute_dtype: float32
|
| 1061 |
+
|
| 1062 |
+
### Framework versions
|
| 1063 |
+
|
| 1064 |
+
|
| 1065 |
- PEFT 0.6.2
|
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 21020682
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6f5b2b89ab2c56eb031a04cbb5ee59066e5bdb3db2ba0f301ae3d11bbd3d80a
|
| 3 |
size 21020682
|