anhnv125/reward-model
Browse files- README.md +4 -24
- model.safetensors +1 -1
- pytorch_model.bin +1 -1
- training_args.bin +1 -1
README.md
CHANGED
@@ -14,7 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
14 |
|
15 |
This model is a fine-tuned version of [ChaiML/reward_models_100_170000000_cp_498032](https://huggingface.co/ChaiML/reward_models_100_170000000_cp_498032) on the None dataset.
|
16 |
It achieves the following results on the evaluation set:
|
17 |
-
- Loss: 0.
|
18 |
|
19 |
## Model description
|
20 |
|
@@ -37,6 +37,8 @@ The following hyperparameters were used during training:
|
|
37 |
- train_batch_size: 16
|
38 |
- eval_batch_size: 16
|
39 |
- seed: 7
|
|
|
|
|
40 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
41 |
- lr_scheduler_type: cosine
|
42 |
- lr_scheduler_warmup_steps: 200
|
@@ -46,29 +48,7 @@ The following hyperparameters were used during training:
|
|
46 |
|
47 |
| Training Loss | Epoch | Step | Validation Loss |
|
48 |
|:-------------:|:-----:|:----:|:---------------:|
|
49 |
-
| 0.
|
50 |
-
| 0.654 | 0.09 | 400 | 0.6472 |
|
51 |
-
| 0.6654 | 0.13 | 600 | 0.6429 |
|
52 |
-
| 0.6381 | 0.17 | 800 | 0.6391 |
|
53 |
-
| 0.6546 | 0.21 | 1000 | 0.6381 |
|
54 |
-
| 0.6575 | 0.26 | 1200 | 0.6352 |
|
55 |
-
| 0.6458 | 0.3 | 1400 | 0.6344 |
|
56 |
-
| 0.6574 | 0.34 | 1600 | 0.6335 |
|
57 |
-
| 0.6409 | 0.39 | 1800 | 0.6314 |
|
58 |
-
| 0.6255 | 0.43 | 2000 | 0.6306 |
|
59 |
-
| 0.6342 | 0.47 | 2200 | 0.6300 |
|
60 |
-
| 0.6117 | 0.51 | 2400 | 0.6298 |
|
61 |
-
| 0.6651 | 0.56 | 2600 | 0.6307 |
|
62 |
-
| 0.6624 | 0.6 | 2800 | 0.6294 |
|
63 |
-
| 0.6213 | 0.64 | 3000 | 0.6286 |
|
64 |
-
| 0.6486 | 0.68 | 3200 | 0.6278 |
|
65 |
-
| 0.6258 | 0.73 | 3400 | 0.6277 |
|
66 |
-
| 0.6336 | 0.77 | 3600 | 0.6278 |
|
67 |
-
| 0.5842 | 0.81 | 3800 | 0.6277 |
|
68 |
-
| 0.6748 | 0.86 | 4000 | 0.6275 |
|
69 |
-
| 0.5985 | 0.9 | 4200 | 0.6275 |
|
70 |
-
| 0.6207 | 0.94 | 4400 | 0.6275 |
|
71 |
-
| 0.5901 | 0.98 | 4600 | 0.6275 |
|
72 |
|
73 |
|
74 |
### Framework versions
|
|
|
14 |
|
15 |
This model is a fine-tuned version of [ChaiML/reward_models_100_170000000_cp_498032](https://huggingface.co/ChaiML/reward_models_100_170000000_cp_498032) on the None dataset.
|
16 |
It achieves the following results on the evaluation set:
|
17 |
+
- Loss: 0.6433
|
18 |
|
19 |
## Model description
|
20 |
|
|
|
37 |
- train_batch_size: 16
|
38 |
- eval_batch_size: 16
|
39 |
- seed: 7
|
40 |
+
- gradient_accumulation_steps: 16
|
41 |
+
- total_train_batch_size: 256
|
42 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
43 |
- lr_scheduler_type: cosine
|
44 |
- lr_scheduler_warmup_steps: 200
|
|
|
48 |
|
49 |
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
+
| 0.6589 | 0.68 | 200 | 0.6433 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
52 |
|
53 |
|
54 |
### Framework versions
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 497780432
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0ce85abb6b8d303da65c45a05e5eec017c2995136a0095956442d1d3f2dec0e3
|
3 |
size 497780432
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 497813601
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b5c5bd47e5abb0ebad25475f8b9eb7fbfbb852d13635dc6ac6160260bb9fc0b1
|
3 |
size 497813601
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4027
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1318d8ccd2e1aedee13b0a06a6be6c81e6da263c655cde47ca1117b3a444f3ab
|
3 |
size 4027
|