Isabelle Cudlitz
commited on
Commit
·
fe5f2b4
1
Parent(s):
a766e08
Upload model
Browse files- README.md +18 -0
- adapter_model.bin +1 -1
README.md
CHANGED
|
@@ -882,4 +882,22 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 882 |
### Framework versions
|
| 883 |
|
| 884 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 885 |
- PEFT 0.6.2
|
|
|
|
| 882 |
### Framework versions
|
| 883 |
|
| 884 |
|
| 885 |
+
- PEFT 0.6.2
|
| 886 |
+
## Training procedure
|
| 887 |
+
|
| 888 |
+
|
| 889 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 890 |
+
- load_in_8bit: True
|
| 891 |
+
- load_in_4bit: False
|
| 892 |
+
- llm_int8_threshold: 6.0
|
| 893 |
+
- llm_int8_skip_modules: None
|
| 894 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 895 |
+
- llm_int8_has_fp16_weight: False
|
| 896 |
+
- bnb_4bit_quant_type: fp4
|
| 897 |
+
- bnb_4bit_use_double_quant: False
|
| 898 |
+
- bnb_4bit_compute_dtype: float32
|
| 899 |
+
|
| 900 |
+
### Framework versions
|
| 901 |
+
|
| 902 |
+
|
| 903 |
- 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 sha256:734894360375255f0bdec7805cf9d816ad0a0cd69562a1bbbe377ab2fa9e6245
|
| 3 |
size 21020682
|