Isabelle Cudlitz
commited on
Commit
·
fa84f67
1
Parent(s):
f217c84
Upload model
Browse files- README.md +18 -0
- adapter_model.bin +1 -1
README.md
CHANGED
|
@@ -486,4 +486,22 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 486 |
### Framework versions
|
| 487 |
|
| 488 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 489 |
- PEFT 0.6.2
|
|
|
|
| 486 |
### Framework versions
|
| 487 |
|
| 488 |
|
| 489 |
+
- PEFT 0.6.2
|
| 490 |
+
## Training procedure
|
| 491 |
+
|
| 492 |
+
|
| 493 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 494 |
+
- load_in_8bit: False
|
| 495 |
+
- load_in_4bit: True
|
| 496 |
+
- llm_int8_threshold: 6.0
|
| 497 |
+
- llm_int8_skip_modules: None
|
| 498 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 499 |
+
- llm_int8_has_fp16_weight: False
|
| 500 |
+
- bnb_4bit_quant_type: fp4
|
| 501 |
+
- bnb_4bit_use_double_quant: True
|
| 502 |
+
- bnb_4bit_compute_dtype: float32
|
| 503 |
+
|
| 504 |
+
### Framework versions
|
| 505 |
+
|
| 506 |
+
|
| 507 |
- 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 sha256:0bcf04d2c6eabc81750f4a9bda3ed5c5c1fba5333ea83d3c81e87878563b6051
|
| 3 |
size 21020682
|