Isabelle Cudlitz
commited on
Commit
·
3b90950
1
Parent(s):
1461ccf
Upload model
Browse files- README.md +19 -0
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -2953,4 +2953,23 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 2953 |
### Framework versions
|
| 2954 |
|
| 2955 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2956 |
- PEFT 0.6.2
|
|
|
|
| 2953 |
### Framework versions
|
| 2954 |
|
| 2955 |
|
| 2956 |
+
- PEFT 0.6.2
|
| 2957 |
+
## Training procedure
|
| 2958 |
+
|
| 2959 |
+
|
| 2960 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 2961 |
+
- quant_method: bitsandbytes
|
| 2962 |
+
- load_in_8bit: True
|
| 2963 |
+
- load_in_4bit: False
|
| 2964 |
+
- llm_int8_threshold: 6.0
|
| 2965 |
+
- llm_int8_skip_modules: None
|
| 2966 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 2967 |
+
- llm_int8_has_fp16_weight: False
|
| 2968 |
+
- bnb_4bit_quant_type: fp4
|
| 2969 |
+
- bnb_4bit_use_double_quant: False
|
| 2970 |
+
- bnb_4bit_compute_dtype: float32
|
| 2971 |
+
|
| 2972 |
+
### Framework versions
|
| 2973 |
+
|
| 2974 |
+
|
| 2975 |
- 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 sha256:346dd635fc3a211a9d1f504d23fe3ee895d08fbc15e1db3e24523e164b3fd635
|
| 3 |
size 20991608
|