Isabelle Cudlitz
commited on
Commit
·
f6868bf
1
Parent(s):
3b90950
Upload model
Browse files- README.md +19 -0
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -2972,4 +2972,23 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 2972 |
### Framework versions
|
| 2973 |
|
| 2974 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2975 |
- PEFT 0.6.2
|
|
|
|
| 2972 |
### Framework versions
|
| 2973 |
|
| 2974 |
|
| 2975 |
+
- PEFT 0.6.2
|
| 2976 |
+
## Training procedure
|
| 2977 |
+
|
| 2978 |
+
|
| 2979 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 2980 |
+
- quant_method: bitsandbytes
|
| 2981 |
+
- load_in_8bit: True
|
| 2982 |
+
- load_in_4bit: False
|
| 2983 |
+
- llm_int8_threshold: 6.0
|
| 2984 |
+
- llm_int8_skip_modules: None
|
| 2985 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 2986 |
+
- llm_int8_has_fp16_weight: False
|
| 2987 |
+
- bnb_4bit_quant_type: fp4
|
| 2988 |
+
- bnb_4bit_use_double_quant: False
|
| 2989 |
+
- bnb_4bit_compute_dtype: float32
|
| 2990 |
+
|
| 2991 |
+
### Framework versions
|
| 2992 |
+
|
| 2993 |
+
|
| 2994 |
- 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 sha256:5f759f2370de8275394a2514bfe184c6f2979d7d12f7a31fc997c84fe68ddb24
|
| 3 |
size 20991608
|