Isabelle Cudlitz
commited on
Commit
·
97c6a04
1
Parent(s):
3cc6bcc
Upload model
Browse files- README.md +19 -0
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -958,4 +958,23 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 958 |
### Framework versions
|
| 959 |
|
| 960 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 961 |
- PEFT 0.6.2
|
|
|
|
| 958 |
### Framework versions
|
| 959 |
|
| 960 |
|
| 961 |
+
- PEFT 0.6.2
|
| 962 |
+
## Training procedure
|
| 963 |
+
|
| 964 |
+
|
| 965 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 966 |
+
- quant_method: bitsandbytes
|
| 967 |
+
- load_in_8bit: True
|
| 968 |
+
- load_in_4bit: False
|
| 969 |
+
- llm_int8_threshold: 6.0
|
| 970 |
+
- llm_int8_skip_modules: None
|
| 971 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 972 |
+
- llm_int8_has_fp16_weight: False
|
| 973 |
+
- bnb_4bit_quant_type: fp4
|
| 974 |
+
- bnb_4bit_use_double_quant: False
|
| 975 |
+
- bnb_4bit_compute_dtype: float32
|
| 976 |
+
|
| 977 |
+
### Framework versions
|
| 978 |
+
|
| 979 |
+
|
| 980 |
- 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 sha256:9d04f5b5863e8f12156c7c35e474ba676aa5de9da14c36c7c309dbcbb72e7334
|
| 3 |
size 20991608
|