Isabelle Cudlitz
commited on
Commit
·
85b1fdd
1
Parent(s):
e4cc87f
Upload model
Browse files- README.md +18 -0
- adapter_model.bin +1 -1
README.md
CHANGED
|
@@ -936,4 +936,22 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 936 |
### Framework versions
|
| 937 |
|
| 938 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 939 |
- PEFT 0.6.2
|
|
|
|
| 936 |
### Framework versions
|
| 937 |
|
| 938 |
|
| 939 |
+
- PEFT 0.6.2
|
| 940 |
+
## Training procedure
|
| 941 |
+
|
| 942 |
+
|
| 943 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 944 |
+
- load_in_8bit: False
|
| 945 |
+
- load_in_4bit: True
|
| 946 |
+
- llm_int8_threshold: 6.0
|
| 947 |
+
- llm_int8_skip_modules: None
|
| 948 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 949 |
+
- llm_int8_has_fp16_weight: False
|
| 950 |
+
- bnb_4bit_quant_type: fp4
|
| 951 |
+
- bnb_4bit_use_double_quant: True
|
| 952 |
+
- bnb_4bit_compute_dtype: float32
|
| 953 |
+
|
| 954 |
+
### Framework versions
|
| 955 |
+
|
| 956 |
+
|
| 957 |
- 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 sha256:9f08a91b9b2cdb51a2a1bdd53778c14b1559633af5233ebae70c1366b769a872
|
| 3 |
size 21020682
|