Isabelle Cudlitz
commited on
Commit
·
54a6669
1
Parent(s):
fa84f67
Upload model
Browse files- README.md +18 -0
- adapter_model.bin +1 -1
README.md
CHANGED
|
@@ -504,4 +504,22 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 504 |
### Framework versions
|
| 505 |
|
| 506 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 507 |
- PEFT 0.6.2
|
|
|
|
| 504 |
### Framework versions
|
| 505 |
|
| 506 |
|
| 507 |
+
- PEFT 0.6.2
|
| 508 |
+
## Training procedure
|
| 509 |
+
|
| 510 |
+
|
| 511 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 512 |
+
- load_in_8bit: False
|
| 513 |
+
- load_in_4bit: True
|
| 514 |
+
- llm_int8_threshold: 6.0
|
| 515 |
+
- llm_int8_skip_modules: None
|
| 516 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 517 |
+
- llm_int8_has_fp16_weight: False
|
| 518 |
+
- bnb_4bit_quant_type: fp4
|
| 519 |
+
- bnb_4bit_use_double_quant: True
|
| 520 |
+
- bnb_4bit_compute_dtype: float32
|
| 521 |
+
|
| 522 |
+
### Framework versions
|
| 523 |
+
|
| 524 |
+
|
| 525 |
- 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 sha256:356cc7523eff3c135d85271ab298fd90b82848ecec2681c638453e40616cae55
|
| 3 |
size 21020682
|