Isabelle Cudlitz
commited on
Commit
·
3cc6bcc
1
Parent(s):
39df573
Upload model
Browse files- README.md +19 -0
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -939,4 +939,23 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 939 |
### Framework versions
|
| 940 |
|
| 941 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 942 |
- PEFT 0.6.2
|
|
|
|
| 939 |
### Framework versions
|
| 940 |
|
| 941 |
|
| 942 |
+
- PEFT 0.6.2
|
| 943 |
+
## Training procedure
|
| 944 |
+
|
| 945 |
+
|
| 946 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 947 |
+
- quant_method: bitsandbytes
|
| 948 |
+
- load_in_8bit: True
|
| 949 |
+
- load_in_4bit: False
|
| 950 |
+
- llm_int8_threshold: 6.0
|
| 951 |
+
- llm_int8_skip_modules: None
|
| 952 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 953 |
+
- llm_int8_has_fp16_weight: False
|
| 954 |
+
- bnb_4bit_quant_type: fp4
|
| 955 |
+
- bnb_4bit_use_double_quant: False
|
| 956 |
+
- bnb_4bit_compute_dtype: float32
|
| 957 |
+
|
| 958 |
+
### Framework versions
|
| 959 |
+
|
| 960 |
+
|
| 961 |
- 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 sha256:3ef184315c2c5d7a7f1a3ed426a10754bc9cb4a2bf190f111a4ae7f3500d6dff
|
| 3 |
size 20991608
|