marcus2000
commited on
Commit
•
4084845
1
Parent(s):
bd83dd9
saiga_demo
Browse files- README.md +4 -12
- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- training_args.bin +1 -1
README.md
CHANGED
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
15 |
|
16 |
This model is a fine-tuned version of [TheBloke/Llama-2-7B-fp16](https://huggingface.co/TheBloke/Llama-2-7B-fp16) on an unknown dataset.
|
17 |
It achieves the following results on the evaluation set:
|
18 |
-
- Loss: 1.
|
19 |
|
20 |
## Model description
|
21 |
|
@@ -42,22 +42,14 @@ The following hyperparameters were used during training:
|
|
42 |
- total_train_batch_size: 20
|
43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
44 |
- lr_scheduler_type: linear
|
45 |
-
- training_steps:
|
46 |
|
47 |
### Training results
|
48 |
|
49 |
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
-
| 1.
|
52 |
-
| 1.
|
53 |
-
| 1.5209 | 4.29 | 30 | 1.6832 |
|
54 |
-
| 1.4711 | 5.71 | 40 | 1.7075 |
|
55 |
-
| 1.4222 | 7.14 | 50 | 1.7207 |
|
56 |
-
| 1.3594 | 8.57 | 60 | 1.7500 |
|
57 |
-
| 1.3276 | 10.0 | 70 | 1.7686 |
|
58 |
-
| 1.2995 | 11.43 | 80 | 1.7832 |
|
59 |
-
| 1.2516 | 12.86 | 90 | 1.7999 |
|
60 |
-
| 1.2647 | 14.29 | 100 | 1.7981 |
|
61 |
|
62 |
|
63 |
### Framework versions
|
|
|
15 |
|
16 |
This model is a fine-tuned version of [TheBloke/Llama-2-7B-fp16](https://huggingface.co/TheBloke/Llama-2-7B-fp16) on an unknown dataset.
|
17 |
It achieves the following results on the evaluation set:
|
18 |
+
- Loss: 1.8757
|
19 |
|
20 |
## Model description
|
21 |
|
|
|
42 |
- total_train_batch_size: 20
|
43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
44 |
- lr_scheduler_type: linear
|
45 |
+
- training_steps: 20
|
46 |
|
47 |
### Training results
|
48 |
|
49 |
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
+
| 1.7956 | 1.43 | 10 | 1.9073 |
|
52 |
+
| 1.7315 | 2.86 | 20 | 1.8757 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
|
54 |
|
55 |
### Framework versions
|
adapter_config.json
CHANGED
@@ -20,10 +20,10 @@
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
23 |
-
"
|
24 |
"k_proj",
|
25 |
-
"
|
26 |
-
"
|
27 |
],
|
28 |
"task_type": "CAUSAL_LM",
|
29 |
"use_dora": false,
|
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
23 |
+
"v_proj",
|
24 |
"k_proj",
|
25 |
+
"q_proj",
|
26 |
+
"o_proj"
|
27 |
],
|
28 |
"task_type": "CAUSAL_LM",
|
29 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 33589040
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:142aa4aadb980f6cd78b8019e923dc97ae7e40161bf41bf0bd3f1861859ee99f
|
3 |
size 33589040
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a04c6a0403c6d4de48a545f873fbfab7b3a71b70eb7de9c9c181cd6b8cdeac96
|
3 |
size 4920
|