Model save
Browse files- README.md +75 -0
- all_results.json +21 -0
- eval_results.json +16 -0
- train_results.json +8 -0
- trainer_state.json +0 -0
README.md
ADDED
@@ -0,0 +1,75 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
library_name: peft
|
4 |
+
tags:
|
5 |
+
- trl
|
6 |
+
- dpo
|
7 |
+
- generated_from_trainer
|
8 |
+
base_model: TomGrc/FusionNet_7Bx2_MoE_v0.1
|
9 |
+
model-index:
|
10 |
+
- name: stealth-finance-v2-dpo-adapter
|
11 |
+
results: []
|
12 |
+
---
|
13 |
+
|
14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
15 |
+
should probably proofread and complete it, then remove this comment. -->
|
16 |
+
|
17 |
+
# stealth-finance-v2-dpo-adapter
|
18 |
+
|
19 |
+
This model is a fine-tuned version of [TomGrc/FusionNet_7Bx2_MoE_v0.1](https://huggingface.co/TomGrc/FusionNet_7Bx2_MoE_v0.1) on the None dataset.
|
20 |
+
It achieves the following results on the evaluation set:
|
21 |
+
- Loss: 0.1290
|
22 |
+
- Rewards/chosen: -0.1799
|
23 |
+
- Rewards/rejected: -6.0696
|
24 |
+
- Rewards/accuracies: 0.8597
|
25 |
+
- Rewards/margins: 5.8897
|
26 |
+
- Logps/rejected: -324.0384
|
27 |
+
- Logps/chosen: -275.3572
|
28 |
+
- Logits/rejected: -0.7749
|
29 |
+
- Logits/chosen: -0.7773
|
30 |
+
|
31 |
+
## Model description
|
32 |
+
|
33 |
+
More information needed
|
34 |
+
|
35 |
+
## Intended uses & limitations
|
36 |
+
|
37 |
+
More information needed
|
38 |
+
|
39 |
+
## Training and evaluation data
|
40 |
+
|
41 |
+
More information needed
|
42 |
+
|
43 |
+
## Training procedure
|
44 |
+
|
45 |
+
### Training hyperparameters
|
46 |
+
|
47 |
+
The following hyperparameters were used during training:
|
48 |
+
- learning_rate: 5e-05
|
49 |
+
- train_batch_size: 1
|
50 |
+
- eval_batch_size: 1
|
51 |
+
- seed: 42
|
52 |
+
- distributed_type: multi-GPU
|
53 |
+
- num_devices: 2
|
54 |
+
- gradient_accumulation_steps: 32
|
55 |
+
- total_train_batch_size: 64
|
56 |
+
- total_eval_batch_size: 2
|
57 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
58 |
+
- lr_scheduler_type: cosine
|
59 |
+
- lr_scheduler_warmup_ratio: 0.1
|
60 |
+
- num_epochs: 1
|
61 |
+
|
62 |
+
### Training results
|
63 |
+
|
64 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
65 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
66 |
+
| 0.3593 | 1.0 | 3280 | 0.1290 | -0.1799 | -6.0696 | 0.8597 | 5.8897 | -324.0384 | -275.3572 | -0.7749 | -0.7773 |
|
67 |
+
|
68 |
+
|
69 |
+
### Framework versions
|
70 |
+
|
71 |
+
- PEFT 0.8.2
|
72 |
+
- Transformers 4.37.2
|
73 |
+
- Pytorch 2.1.2+cu121
|
74 |
+
- Datasets 2.14.6
|
75 |
+
- Tokenizers 0.15.0
|
all_results.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_logits/chosen": -0.7773212790489197,
|
4 |
+
"eval_logits/rejected": -0.7749085426330566,
|
5 |
+
"eval_logps/chosen": -275.3572082519531,
|
6 |
+
"eval_logps/rejected": -324.0383605957031,
|
7 |
+
"eval_loss": 0.12900209426879883,
|
8 |
+
"eval_rewards/accuracies": 0.8596742749214172,
|
9 |
+
"eval_rewards/chosen": -0.17991267144680023,
|
10 |
+
"eval_rewards/margins": 5.889674186706543,
|
11 |
+
"eval_rewards/rejected": -6.069586277008057,
|
12 |
+
"eval_runtime": 76159.9506,
|
13 |
+
"eval_samples": 197968,
|
14 |
+
"eval_samples_per_second": 2.599,
|
15 |
+
"eval_steps_per_second": 1.3,
|
16 |
+
"train_loss": 0.3931771684165408,
|
17 |
+
"train_runtime": 248473.0607,
|
18 |
+
"train_samples": 209976,
|
19 |
+
"train_samples_per_second": 0.845,
|
20 |
+
"train_steps_per_second": 0.013
|
21 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_logits/chosen": -0.7773212790489197,
|
4 |
+
"eval_logits/rejected": -0.7749085426330566,
|
5 |
+
"eval_logps/chosen": -275.3572082519531,
|
6 |
+
"eval_logps/rejected": -324.0383605957031,
|
7 |
+
"eval_loss": 0.12900209426879883,
|
8 |
+
"eval_rewards/accuracies": 0.8596742749214172,
|
9 |
+
"eval_rewards/chosen": -0.17991267144680023,
|
10 |
+
"eval_rewards/margins": 5.889674186706543,
|
11 |
+
"eval_rewards/rejected": -6.069586277008057,
|
12 |
+
"eval_runtime": 76159.9506,
|
13 |
+
"eval_samples": 197968,
|
14 |
+
"eval_samples_per_second": 2.599,
|
15 |
+
"eval_steps_per_second": 1.3
|
16 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"train_loss": 0.3931771684165408,
|
4 |
+
"train_runtime": 248473.0607,
|
5 |
+
"train_samples": 209976,
|
6 |
+
"train_samples_per_second": 0.845,
|
7 |
+
"train_steps_per_second": 0.013
|
8 |
+
}
|
trainer_state.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|