End of training
Browse files
README.md
ADDED
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model: meta-llama/Llama-2-7b-hf
|
3 |
+
library_name: peft
|
4 |
+
license: llama2
|
5 |
+
tags:
|
6 |
+
- trl
|
7 |
+
- dpo
|
8 |
+
- generated_from_trainer
|
9 |
+
model-index:
|
10 |
+
- name: Llama-2-7b-hf-DPO-LookAhead-5_TTree1.4_TT0.9_TP0.7_TE0.2_V4
|
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 |
+
# Llama-2-7b-hf-DPO-LookAhead-5_TTree1.4_TT0.9_TP0.7_TE0.2_V4
|
18 |
+
|
19 |
+
This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on the None dataset.
|
20 |
+
It achieves the following results on the evaluation set:
|
21 |
+
- Loss: 1.2125
|
22 |
+
- Rewards/chosen: -3.3104
|
23 |
+
- Rewards/rejected: -2.9319
|
24 |
+
- Rewards/accuracies: 0.4167
|
25 |
+
- Rewards/margins: -0.3786
|
26 |
+
- Logps/rejected: -192.9225
|
27 |
+
- Logps/chosen: -170.2794
|
28 |
+
- Logits/rejected: 0.1199
|
29 |
+
- Logits/chosen: 0.1595
|
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: 2
|
50 |
+
- eval_batch_size: 2
|
51 |
+
- seed: 42
|
52 |
+
- gradient_accumulation_steps: 2
|
53 |
+
- total_train_batch_size: 4
|
54 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
55 |
+
- lr_scheduler_type: cosine
|
56 |
+
- lr_scheduler_warmup_steps: 10
|
57 |
+
- num_epochs: 3
|
58 |
+
|
59 |
+
### Training results
|
60 |
+
|
61 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
62 |
+
|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
63 |
+
| 0.6179 | 0.3027 | 79 | 0.7115 | -0.1031 | -0.0593 | 0.25 | -0.0438 | -164.1966 | -138.2057 | 0.5429 | 0.5748 |
|
64 |
+
| 0.6065 | 0.6054 | 158 | 0.7348 | -0.0751 | 0.0129 | 0.25 | -0.0879 | -163.4753 | -137.9259 | 0.5242 | 0.5565 |
|
65 |
+
| 0.621 | 0.9080 | 237 | 0.7932 | -0.0433 | 0.1366 | 0.5 | -0.1800 | -162.2375 | -137.6083 | 0.4932 | 0.5259 |
|
66 |
+
| 0.4714 | 1.2107 | 316 | 0.7928 | -0.6963 | -0.5927 | 0.5 | -0.1037 | -169.5308 | -144.1387 | 0.4698 | 0.5037 |
|
67 |
+
| 0.3829 | 1.5134 | 395 | 0.8637 | -1.6604 | -1.5528 | 0.3333 | -0.1075 | -179.1323 | -153.7787 | 0.3664 | 0.4026 |
|
68 |
+
| 0.3589 | 1.8161 | 474 | 0.9222 | -1.4397 | -1.1360 | 0.25 | -0.3037 | -174.9637 | -151.5720 | 0.3400 | 0.3770 |
|
69 |
+
| 0.2138 | 2.1188 | 553 | 0.9860 | -1.9991 | -1.6486 | 0.3333 | -0.3505 | -180.0903 | -157.1666 | 0.2605 | 0.2992 |
|
70 |
+
| 0.0437 | 2.4215 | 632 | 1.1781 | -3.1628 | -2.7961 | 0.4167 | -0.3666 | -191.5652 | -168.8030 | 0.1441 | 0.1838 |
|
71 |
+
| 0.1667 | 2.7241 | 711 | 1.2125 | -3.3104 | -2.9319 | 0.4167 | -0.3786 | -192.9225 | -170.2794 | 0.1199 | 0.1595 |
|
72 |
+
|
73 |
+
|
74 |
+
### Framework versions
|
75 |
+
|
76 |
+
- PEFT 0.12.0
|
77 |
+
- Transformers 4.44.0
|
78 |
+
- Pytorch 2.4.0+cu121
|
79 |
+
- Datasets 3.1.0
|
80 |
+
- Tokenizers 0.19.1
|