NicholasCorrado
commited on
Commit
•
0283f96
1
Parent(s):
39e6203
Model save
Browse files- README.md +100 -0
- all_results.json +9 -0
- generation_config.json +6 -0
- train_results.json +9 -0
README.md
ADDED
@@ -0,0 +1,100 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: transformers
|
3 |
+
license: apache-2.0
|
4 |
+
base_model: alignment-handbook/zephyr-7b-sft-full
|
5 |
+
tags:
|
6 |
+
- trl
|
7 |
+
- dpo
|
8 |
+
- generated_from_trainer
|
9 |
+
model-index:
|
10 |
+
- name: zephyr-7b-uf-rlced-conifer-group-dpo-2e-alr-0.1
|
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 |
+
# zephyr-7b-uf-rlced-conifer-group-dpo-2e-alr-0.1
|
18 |
+
|
19 |
+
This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on an unknown dataset.
|
20 |
+
It achieves the following results on the evaluation set:
|
21 |
+
- Loss: 0.2391
|
22 |
+
- Rewards/chosen: -3.1721
|
23 |
+
- Rewards/rejected: -8.7679
|
24 |
+
- Rewards/accuracies: 0.8788
|
25 |
+
- Rewards/margins: 5.5958
|
26 |
+
- Logps/rejected: -1280.5232
|
27 |
+
- Logps/chosen: -709.6791
|
28 |
+
- Logits/rejected: 2.9862
|
29 |
+
- Logits/chosen: 0.4871
|
30 |
+
- Excess Loss: 0.0302
|
31 |
+
- Alpha 0 Uf: 0.2677
|
32 |
+
- Alpha 1 Rlced Conifer: 0.7323
|
33 |
+
- Rewards/chosen 1 Rlced Conifer: -3.3519
|
34 |
+
- Rewards/rejected 1 Rlced Conifer: -10.1355
|
35 |
+
- Rewards/accuracies 1 Rlced Conifer: 0.9088
|
36 |
+
- Rewards/margins 1 Rlced Conifer: 6.7836
|
37 |
+
- Logps/rejected 1 Rlced Conifer: -1461.0847
|
38 |
+
- Logps/chosen 1 Rlced Conifer: -758.7692
|
39 |
+
- Logits/rejected 1 Rlced Conifer: 2.9834
|
40 |
+
- Logits/chosen 1 Rlced Conifer: 0.2872
|
41 |
+
- Task Loss 1 Rlced Conifer: 0.1744
|
42 |
+
- Task Excess Loss 1 Rlced Conifer: 0.0378
|
43 |
+
- Rewards/chosen 0 Uf: -2.5137
|
44 |
+
- Rewards/rejected 0 Uf: -3.9578
|
45 |
+
- Rewards/accuracies 0 Uf: 0.7751
|
46 |
+
- Rewards/margins 0 Uf: 1.4442
|
47 |
+
- Logps/rejected 0 Uf: -637.3895
|
48 |
+
- Logps/chosen 0 Uf: -540.6270
|
49 |
+
- Logits/rejected 0 Uf: 3.2024
|
50 |
+
- Logits/chosen 0 Uf: 1.0821
|
51 |
+
- Task Loss 0 Uf: 0.5033
|
52 |
+
- Task Excess Loss 0 Uf: 0.0690
|
53 |
+
|
54 |
+
## Model description
|
55 |
+
|
56 |
+
More information needed
|
57 |
+
|
58 |
+
## Intended uses & limitations
|
59 |
+
|
60 |
+
More information needed
|
61 |
+
|
62 |
+
## Training and evaluation data
|
63 |
+
|
64 |
+
More information needed
|
65 |
+
|
66 |
+
## Training procedure
|
67 |
+
|
68 |
+
### Training hyperparameters
|
69 |
+
|
70 |
+
The following hyperparameters were used during training:
|
71 |
+
- learning_rate: 5e-07
|
72 |
+
- train_batch_size: 8
|
73 |
+
- eval_batch_size: 8
|
74 |
+
- seed: 42
|
75 |
+
- distributed_type: multi-GPU
|
76 |
+
- num_devices: 8
|
77 |
+
- gradient_accumulation_steps: 4
|
78 |
+
- total_train_batch_size: 256
|
79 |
+
- total_eval_batch_size: 64
|
80 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
81 |
+
- lr_scheduler_type: cosine
|
82 |
+
- lr_scheduler_warmup_ratio: 0.1
|
83 |
+
- num_epochs: 2
|
84 |
+
|
85 |
+
### Training results
|
86 |
+
|
87 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | Excess Loss | Alpha 0 Uf | Alpha 1 Rlced Conifer | Rewards/chosen 1 Rlced Conifer | Rewards/rejected 1 Rlced Conifer | Rewards/accuracies 1 Rlced Conifer | Rewards/margins 1 Rlced Conifer | Logps/rejected 1 Rlced Conifer | Logps/chosen 1 Rlced Conifer | Logits/rejected 1 Rlced Conifer | Logits/chosen 1 Rlced Conifer | Task Loss 1 Rlced Conifer | Task Excess Loss 1 Rlced Conifer | Rewards/chosen 0 Uf | Rewards/rejected 0 Uf | Rewards/accuracies 0 Uf | Rewards/margins 0 Uf | Logps/rejected 0 Uf | Logps/chosen 0 Uf | Logits/rejected 0 Uf | Logits/chosen 0 Uf | Task Loss 0 Uf | Task Excess Loss 0 Uf |
|
88 |
+
|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:-----------:|:----------:|:---------------------:|:------------------------------:|:--------------------------------:|:----------------------------------:|:-------------------------------:|:------------------------------:|:----------------------------:|:-------------------------------:|:-----------------------------:|:-------------------------:|:--------------------------------:|:-------------------:|:---------------------:|:-----------------------:|:--------------------:|:-------------------:|:-----------------:|:--------------------:|:------------------:|:--------------:|:---------------------:|
|
89 |
+
| 0.1882 | 0.4997 | 360 | 0.2996 | -1.6886 | -4.1417 | 0.8609 | 2.4532 | -817.9084 | -561.3260 | 1.4584 | 0.3084 | 0.0858 | 0.8164 | 0.1836 | -1.7447 | -4.6283 | 0.8926 | 2.8836 | -910.3677 | -598.0539 | 1.2471 | 0.1366 | 0.2441 | 0.1077 | -1.4688 | -2.4264 | 0.7375 | 0.9576 | -484.2446 | -436.1386 | 2.3554 | 0.8449 | 0.5159 | 0.0745 |
|
90 |
+
| 0.1534 | 0.9993 | 720 | 0.2788 | -1.6895 | -4.6113 | 0.8656 | 2.9218 | -864.8680 | -561.4199 | 1.5835 | 0.1282 | 0.0703 | 0.8639 | 0.1361 | -1.7298 | -5.1653 | 0.8921 | 3.4355 | -964.0696 | -596.5645 | 1.3282 | -0.0899 | 0.2304 | 0.0945 | -1.5189 | -2.6316 | 0.7670 | 1.1128 | -504.7690 | -441.1461 | 2.6475 | 0.8112 | 0.4886 | 0.0496 |
|
91 |
+
| 0.0947 | 1.4990 | 1080 | 0.2421 | -2.6372 | -7.6503 | 0.8797 | 5.0132 | -1168.7697 | -656.1883 | 2.9592 | 0.5518 | 0.0336 | 0.2372 | 0.7628 | -2.7432 | -8.7916 | 0.9108 | 6.0484 | -1326.6932 | -697.9009 | 2.9155 | 0.3378 | 0.1806 | 0.0448 | -2.2397 | -3.6057 | 0.7721 | 1.3660 | -602.1759 | -513.2244 | 3.3160 | 1.1969 | 0.4985 | 0.0623 |
|
92 |
+
| 0.0894 | 1.9986 | 1440 | 0.2391 | -3.1721 | -8.7679 | 0.8788 | 5.5958 | -1280.5232 | -709.6791 | 2.9862 | 0.4871 | 0.0302 | 0.2677 | 0.7323 | -3.3519 | -10.1355 | 0.9088 | 6.7836 | -1461.0847 | -758.7692 | 2.9834 | 0.2872 | 0.1744 | 0.0378 | -2.5137 | -3.9578 | 0.7751 | 1.4442 | -637.3895 | -540.6270 | 3.2024 | 1.0821 | 0.5033 | 0.0690 |
|
93 |
+
|
94 |
+
|
95 |
+
### Framework versions
|
96 |
+
|
97 |
+
- Transformers 4.44.2
|
98 |
+
- Pytorch 2.2.0a0+81ea7a4
|
99 |
+
- Datasets 2.21.0
|
100 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.9986120749479528,
|
3 |
+
"total_flos": 0.0,
|
4 |
+
"train_loss": 0.1639749237232738,
|
5 |
+
"train_runtime": 41844.7516,
|
6 |
+
"train_samples": 184443,
|
7 |
+
"train_samples_per_second": 8.816,
|
8 |
+
"train_steps_per_second": 0.034
|
9 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 1,
|
4 |
+
"eos_token_id": 2,
|
5 |
+
"transformers_version": "4.44.2"
|
6 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 1.9986120749479528,
|
3 |
+
"total_flos": 0.0,
|
4 |
+
"train_loss": 0.1639749237232738,
|
5 |
+
"train_runtime": 41844.7516,
|
6 |
+
"train_samples": 184443,
|
7 |
+
"train_samples_per_second": 8.816,
|
8 |
+
"train_steps_per_second": 0.034
|
9 |
+
}
|