KeerthiPriya
commited on
Commit
•
b32aba2
1
Parent(s):
df108a3
End of training
Browse files
README.md
ADDED
@@ -0,0 +1,84 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
library_name: peft
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
base_model: filipealmeida/Mistral-7B-Instruct-v0.1-sharded
|
7 |
+
model-index:
|
8 |
+
- name: mistral7b-finetune-20k-withnoclass
|
9 |
+
results: []
|
10 |
+
---
|
11 |
+
|
12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
13 |
+
should probably proofread and complete it, then remove this comment. -->
|
14 |
+
|
15 |
+
# mistral7b-finetune-20k-withnoclass
|
16 |
+
|
17 |
+
This model is a fine-tuned version of [filipealmeida/Mistral-7B-Instruct-v0.1-sharded](https://huggingface.co/filipealmeida/Mistral-7B-Instruct-v0.1-sharded) on the None dataset.
|
18 |
+
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 0.9941
|
20 |
+
|
21 |
+
## Model description
|
22 |
+
|
23 |
+
More information needed
|
24 |
+
|
25 |
+
## Intended uses & limitations
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Training and evaluation data
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training procedure
|
34 |
+
|
35 |
+
### Training hyperparameters
|
36 |
+
|
37 |
+
The following hyperparameters were used during training:
|
38 |
+
- learning_rate: 0.0002
|
39 |
+
- train_batch_size: 8
|
40 |
+
- eval_batch_size: 8
|
41 |
+
- seed: 42
|
42 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
43 |
+
- lr_scheduler_type: cosine
|
44 |
+
- training_steps: 2500
|
45 |
+
- mixed_precision_training: Native AMP
|
46 |
+
|
47 |
+
### Training results
|
48 |
+
|
49 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
+
| 1.8094 | 0.04 | 100 | 1.4332 |
|
52 |
+
| 1.274 | 0.08 | 200 | 1.2291 |
|
53 |
+
| 1.1812 | 0.12 | 300 | 1.1865 |
|
54 |
+
| 1.1474 | 0.16 | 400 | 1.1542 |
|
55 |
+
| 1.1421 | 0.2 | 500 | 1.1370 |
|
56 |
+
| 1.101 | 0.24 | 600 | 1.1154 |
|
57 |
+
| 1.1111 | 0.29 | 700 | 1.0999 |
|
58 |
+
| 1.0778 | 0.33 | 800 | 1.0841 |
|
59 |
+
| 1.0421 | 0.37 | 900 | 1.0784 |
|
60 |
+
| 1.0468 | 0.41 | 1000 | 1.0644 |
|
61 |
+
| 1.0389 | 0.45 | 1100 | 1.0540 |
|
62 |
+
| 1.0023 | 0.49 | 1200 | 1.0434 |
|
63 |
+
| 1.0396 | 0.53 | 1300 | 1.0342 |
|
64 |
+
| 0.9991 | 0.57 | 1400 | 1.0260 |
|
65 |
+
| 1.0304 | 0.61 | 1500 | 1.0238 |
|
66 |
+
| 1.0033 | 0.65 | 1600 | 1.0159 |
|
67 |
+
| 1.0065 | 0.69 | 1700 | 1.0109 |
|
68 |
+
| 0.9587 | 0.73 | 1800 | 1.0072 |
|
69 |
+
| 0.9725 | 0.78 | 1900 | 1.0025 |
|
70 |
+
| 0.9738 | 0.82 | 2000 | 0.9997 |
|
71 |
+
| 0.9816 | 0.86 | 2100 | 0.9972 |
|
72 |
+
| 0.9858 | 0.9 | 2200 | 0.9956 |
|
73 |
+
| 0.9477 | 0.94 | 2300 | 0.9946 |
|
74 |
+
| 0.9834 | 0.98 | 2400 | 0.9941 |
|
75 |
+
| 0.943 | 1.02 | 2500 | 0.9941 |
|
76 |
+
|
77 |
+
|
78 |
+
### Framework versions
|
79 |
+
|
80 |
+
- PEFT 0.7.1
|
81 |
+
- Transformers 4.37.0.dev0
|
82 |
+
- Pytorch 2.0.0
|
83 |
+
- Datasets 2.1.0
|
84 |
+
- Tokenizers 0.15.0
|