Add readme
Browse files
README.md
ADDED
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: llama2
|
3 |
+
base_model: codellama/CodeLlama-7b-Instruct-hf
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
metrics:
|
7 |
+
- accuracy
|
8 |
+
- bleu
|
9 |
+
- sacrebleu
|
10 |
+
- rouge
|
11 |
+
model-index:
|
12 |
+
- name: CodeLlama-7b-Instruct-hf_Fi__components_size_252_epochs_10_2024-06-21_09-35-27_3556547
|
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 |
+
# CodeLlama-7b-Instruct-hf_Fi__components_size_252_epochs_10_2024-06-21_09-35-27_3556547
|
20 |
+
|
21 |
+
This model is a fine-tuned version of [codellama/CodeLlama-7b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-7b-Instruct-hf) on the None dataset.
|
22 |
+
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 1.9096
|
24 |
+
- Accuracy: 0.462
|
25 |
+
- Chrf: 0.297
|
26 |
+
- Bleu: 0.225
|
27 |
+
- Sacrebleu: 0.2
|
28 |
+
- Rouge1: 0.472
|
29 |
+
- Rouge2: 0.3
|
30 |
+
- Rougel: 0.459
|
31 |
+
- Rougelsum: 0.471
|
32 |
+
- Meteor: 0.505
|
33 |
+
|
34 |
+
## Model description
|
35 |
+
|
36 |
+
More information needed
|
37 |
+
|
38 |
+
## Intended uses & limitations
|
39 |
+
|
40 |
+
More information needed
|
41 |
+
|
42 |
+
## Training and evaluation data
|
43 |
+
|
44 |
+
More information needed
|
45 |
+
|
46 |
+
## Training procedure
|
47 |
+
|
48 |
+
### Training hyperparameters
|
49 |
+
|
50 |
+
The following hyperparameters were used during training:
|
51 |
+
- learning_rate: 0.001
|
52 |
+
- train_batch_size: 1
|
53 |
+
- eval_batch_size: 1
|
54 |
+
- seed: 3407
|
55 |
+
- distributed_type: multi-GPU
|
56 |
+
- num_devices: 4
|
57 |
+
- total_train_batch_size: 4
|
58 |
+
- total_eval_batch_size: 4
|
59 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-06
|
60 |
+
- lr_scheduler_type: linear
|
61 |
+
- lr_scheduler_warmup_steps: 252
|
62 |
+
- training_steps: 2520
|
63 |
+
|
64 |
+
### Training results
|
65 |
+
|
66 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Chrf | Bleu | Sacrebleu | Rouge1 | Rouge2 | Rougel | Rougelsum | Meteor |
|
67 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|:-----:|:-----:|:---------:|:------:|:------:|:------:|:---------:|:------:|
|
68 |
+
| 0.063 | 4.0 | 252 | 3.6864 | 0.457 | 0.044 | 0.0 | 0.0 | 0.044 | 0.0 | 0.03 | 0.03 | 0.138 |
|
69 |
+
| 0.0742 | 8.0 | 504 | 2.7260 | 0.474 | 0.104 | 0.036 | 0.0 | 0.148 | 0.009 | 0.126 | 0.143 | 0.24 |
|
70 |
+
| 0.0774 | 12.0 | 756 | 2.6054 | 0.461 | 0.159 | 0.099 | 0.1 | 0.315 | 0.149 | 0.306 | 0.308 | 0.325 |
|
71 |
+
| 0.7995 | 16.0 | 1008 | 2.4395 | 0.465 | 0.215 | 0.119 | 0.1 | 0.393 | 0.178 | 0.365 | 0.379 | 0.359 |
|
72 |
+
| 0.1761 | 20.0 | 1260 | 2.4190 | 0.482 | 0.249 | 0.164 | 0.2 | 0.356 | 0.194 | 0.34 | 0.355 | 0.39 |
|
73 |
+
| 0.4002 | 24.0 | 1512 | 2.1404 | 0.462 | 0.251 | 0.188 | 0.2 | 0.418 | 0.269 | 0.4 | 0.409 | 0.437 |
|
74 |
+
| 0.0254 | 28.0 | 1764 | 2.0202 | 0.46 | 0.295 | 0.192 | 0.2 | 0.484 | 0.308 | 0.461 | 0.478 | 0.463 |
|
75 |
+
| 0.1469 | 32.0 | 2016 | 1.9957 | 0.462 | 0.289 | 0.225 | 0.2 | 0.448 | 0.291 | 0.44 | 0.443 | 0.482 |
|
76 |
+
| 0.0346 | 36.0 | 2268 | 1.9562 | 0.46 | 0.293 | 0.2 | 0.2 | 0.474 | 0.278 | 0.452 | 0.471 | 0.491 |
|
77 |
+
| 0.0378 | 40.0 | 2520 | 1.9096 | 0.462 | 0.297 | 0.225 | 0.2 | 0.472 | 0.3 | 0.459 | 0.471 | 0.505 |
|
78 |
+
|
79 |
+
|
80 |
+
### Framework versions
|
81 |
+
|
82 |
+
- Transformers 4.37.0
|
83 |
+
- Pytorch 2.2.1+cu121
|
84 |
+
- Datasets 2.20.0
|
85 |
+
- Tokenizers 0.15.2
|