Commit
•
71352a4
1
Parent(s):
19893c9
Model save
Browse files
README.md
ADDED
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: other
|
3 |
+
library_name: peft
|
4 |
+
tags:
|
5 |
+
- trl
|
6 |
+
- sft
|
7 |
+
- generated_from_trainer
|
8 |
+
base_model: NousResearch/Meta-Llama-3-8B-Instruct
|
9 |
+
datasets:
|
10 |
+
- generator
|
11 |
+
model-index:
|
12 |
+
- name: clearpolicy-llama-3-8binstruct-v4
|
13 |
+
results: []
|
14 |
+
---
|
15 |
+
|
16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
+
should probably proofread and complete it, then remove this comment. -->
|
18 |
+
|
19 |
+
# clearpolicy-llama-3-8binstruct-v4
|
20 |
+
|
21 |
+
This model is a fine-tuned version of [NousResearch/Meta-Llama-3-8B-Instruct](https://huggingface.co/NousResearch/Meta-Llama-3-8B-Instruct) on the generator dataset.
|
22 |
+
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 0.7051
|
24 |
+
|
25 |
+
## Model description
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Intended uses & limitations
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training and evaluation data
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training procedure
|
38 |
+
|
39 |
+
### Training hyperparameters
|
40 |
+
|
41 |
+
The following hyperparameters were used during training:
|
42 |
+
- learning_rate: 0.0002
|
43 |
+
- train_batch_size: 2
|
44 |
+
- eval_batch_size: 8
|
45 |
+
- seed: 42
|
46 |
+
- gradient_accumulation_steps: 2
|
47 |
+
- total_train_batch_size: 4
|
48 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
49 |
+
- lr_scheduler_type: constant
|
50 |
+
- lr_scheduler_warmup_ratio: 0.03
|
51 |
+
- num_epochs: 4
|
52 |
+
- mixed_precision_training: Native AMP
|
53 |
+
|
54 |
+
### Training results
|
55 |
+
|
56 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
57 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
58 |
+
| No log | 1.0 | 7 | 1.2237 |
|
59 |
+
| No log | 2.0 | 14 | 0.8634 |
|
60 |
+
| No log | 3.0 | 21 | 0.7050 |
|
61 |
+
| No log | 4.0 | 28 | 0.7051 |
|
62 |
+
|
63 |
+
|
64 |
+
### Framework versions
|
65 |
+
|
66 |
+
- PEFT 0.11.0
|
67 |
+
- Transformers 4.40.2
|
68 |
+
- Pytorch 2.2.1+cu121
|
69 |
+
- Datasets 2.19.1
|
70 |
+
- Tokenizers 0.19.1
|