Isabelle Cudlitz commited on
Commit ·
9163f5b
1
Parent(s): 35d0c6b
Upload model
Browse files- README.md +19 -0
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -2877,4 +2877,23 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 2877 |
### Framework versions
|
| 2878 |
|
| 2879 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2880 |
- PEFT 0.6.2
|
|
|
|
| 2877 |
### Framework versions
|
| 2878 |
|
| 2879 |
|
| 2880 |
+
- PEFT 0.6.2
|
| 2881 |
+
## Training procedure
|
| 2882 |
+
|
| 2883 |
+
|
| 2884 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 2885 |
+
- quant_method: bitsandbytes
|
| 2886 |
+
- load_in_8bit: True
|
| 2887 |
+
- load_in_4bit: False
|
| 2888 |
+
- llm_int8_threshold: 6.0
|
| 2889 |
+
- llm_int8_skip_modules: None
|
| 2890 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 2891 |
+
- llm_int8_has_fp16_weight: False
|
| 2892 |
+
- bnb_4bit_quant_type: fp4
|
| 2893 |
+
- bnb_4bit_use_double_quant: False
|
| 2894 |
+
- bnb_4bit_compute_dtype: float32
|
| 2895 |
+
|
| 2896 |
+
### Framework versions
|
| 2897 |
+
|
| 2898 |
+
|
| 2899 |
- PEFT 0.6.2
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 20991608
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b17672cb787e3440ea1e16080f3acdc4d8b5dda64bad0dd5ef3e7211a2ced3a7
|
| 3 |
size 20991608
|