Commit
•
c12994f
1
Parent(s):
df83e6b
Upload model
Browse files- README.md +26 -0
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -280,4 +280,30 @@ The following `bitsandbytes` quantization config was used during training:
|
|
| 280 |
### Framework versions
|
| 281 |
|
| 282 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 283 |
- PEFT 0.6.0.dev0
|
|
|
|
| 280 |
### Framework versions
|
| 281 |
|
| 282 |
|
| 283 |
+
- PEFT 0.6.0.dev0
|
| 284 |
+
## Training procedure
|
| 285 |
+
|
| 286 |
+
|
| 287 |
+
The following `bitsandbytes` quantization config was used during training:
|
| 288 |
+
- quant_method: QuantizationMethod.BITS_AND_BYTES
|
| 289 |
+
- load_in_8bit: True
|
| 290 |
+
- load_in_4bit: False
|
| 291 |
+
- llm_int8_threshold: 6.0
|
| 292 |
+
- llm_int8_skip_modules: ['decoder', 'lm_head', 'wo']
|
| 293 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
| 294 |
+
- llm_int8_has_fp16_weight: False
|
| 295 |
+
- bnb_4bit_quant_type: fp4
|
| 296 |
+
- bnb_4bit_use_double_quant: False
|
| 297 |
+
- bnb_4bit_compute_dtype: float32
|
| 298 |
+
|
| 299 |
+
### Framework versions
|
| 300 |
+
|
| 301 |
+
|
| 302 |
+
- PEFT 0.6.0.dev0
|
| 303 |
+
## Training procedure
|
| 304 |
+
|
| 305 |
+
|
| 306 |
+
### Framework versions
|
| 307 |
+
|
| 308 |
+
|
| 309 |
- PEFT 0.6.0.dev0
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 218138576
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dca74f2290e15a82e4f0a6650be2399f1f1ab8848af43165bcc06d4a05b37a31
|
| 3 |
size 218138576
|