Isabelle Cudlitz
commited on
Commit
·
fd8d2e2
1
Parent(s):
72ca0b7
Upload model
Browse files- README.md +18 -0
- adapter_model.bin +1 -1
README.md
CHANGED
|
@@ -1512,4 +1512,22 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 1512 |
### Framework versions
|
| 1513 |
|
| 1514 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1515 |
- PEFT 0.6.2
|
|
|
|
| 1512 |
### Framework versions
|
| 1513 |
|
| 1514 |
|
| 1515 |
+
- PEFT 0.6.2
|
| 1516 |
+
## Training procedure
|
| 1517 |
+
|
| 1518 |
+
|
| 1519 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 1520 |
+
- load_in_8bit: False
|
| 1521 |
+
- load_in_4bit: True
|
| 1522 |
+
- llm_int8_threshold: 6.0
|
| 1523 |
+
- llm_int8_skip_modules: None
|
| 1524 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 1525 |
+
- llm_int8_has_fp16_weight: False
|
| 1526 |
+
- bnb_4bit_quant_type: fp4
|
| 1527 |
+
- bnb_4bit_use_double_quant: True
|
| 1528 |
+
- bnb_4bit_compute_dtype: float32
|
| 1529 |
+
|
| 1530 |
+
### Framework versions
|
| 1531 |
+
|
| 1532 |
+
|
| 1533 |
- 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 sha256:3f8cb7721fbf71a36406d263bf040777a8df33a152c2fb17ad05dbda80cbc218
|
| 3 |
size 21020682
|