Isabelle Cudlitz
commited on
Commit
·
c5f66ae
1
Parent(s):
f5fb524
Upload model
Browse files- README.md +19 -0
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -3029,4 +3029,23 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 3029 |
### Framework versions
|
| 3030 |
|
| 3031 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3032 |
- PEFT 0.6.2
|
|
|
|
| 3029 |
### Framework versions
|
| 3030 |
|
| 3031 |
|
| 3032 |
+
- PEFT 0.6.2
|
| 3033 |
+
## Training procedure
|
| 3034 |
+
|
| 3035 |
+
|
| 3036 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 3037 |
+
- quant_method: bitsandbytes
|
| 3038 |
+
- load_in_8bit: True
|
| 3039 |
+
- load_in_4bit: False
|
| 3040 |
+
- llm_int8_threshold: 6.0
|
| 3041 |
+
- llm_int8_skip_modules: None
|
| 3042 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 3043 |
+
- llm_int8_has_fp16_weight: False
|
| 3044 |
+
- bnb_4bit_quant_type: fp4
|
| 3045 |
+
- bnb_4bit_use_double_quant: False
|
| 3046 |
+
- bnb_4bit_compute_dtype: float32
|
| 3047 |
+
|
| 3048 |
+
### Framework versions
|
| 3049 |
+
|
| 3050 |
+
|
| 3051 |
- 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:f52400319ffe0cdfcac14a1da33e8bf3c852aa78e81e25d7f7cffe3678e45ad5
|
| 3 |
size 20991608
|