Update README.md
Browse files
README.md
CHANGED
@@ -24,3 +24,109 @@ configs:
|
|
24 |
- split: test
|
25 |
path: data/test-*
|
26 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
24 |
- split: test
|
25 |
path: data/test-*
|
26 |
---
|
27 |
+
|
28 |
+
## Explanation
|
29 |
+
A small experiment insipred by the [Mistral playing DOOM experiment](https://github.com/umuthopeyildirim/DOOM-Mistral/tree/main) from the Mistral Hackathon
|
30 |
+
|
31 |
+
**How it works?**
|
32 |
+
```
|
33 |
+
Audio -> Waveform Visualization -> Waveform ASCII Art -> Finetune Mistral on ASCII Art to predict text from ASCII Art
|
34 |
+
```
|
35 |
+
|
36 |
+
**Quick video explanation**
|
37 |
+
<video width="820" controls src="https://cdn-uploads.huggingface.co/production/uploads/6382255fcae34727b9cc149e/nCB8Qu8QwDbJAKcq9IzPE.mp4"></video>
|
38 |
+
|
39 |
+
|
40 |
+
#### Example Waveform
|
41 |
+
![image/png](https://cdn-uploads.huggingface.co/production/uploads/6382255fcae34727b9cc149e/4ui-5eJihKm0vYQ33MRfZ.png)
|
42 |
+
|
43 |
+
|
44 |
+
#### Example ASCII Art
|
45 |
+
```
|
46 |
+
|
47 |
+
y
|
48 |
+
x
|
49 |
+
x
|
50 |
+
x
|
51 |
+
ux
|
52 |
+
tx
|
53 |
+
dx
|
54 |
+
9 ex v
|
55 |
+
4 dr v
|
56 |
+
4 do s
|
57 |
+
1Y 6o e o5
|
58 |
+
x 1Y y 6o ze 7 iYY
|
59 |
+
e YYz r Ys Ym xd Y gYY
|
60 |
+
6 YYY r Ys Yl xd sY Y eYY Y
|
61 |
+
z 6 YYY y s k z Ys Yl xd sY Y m vYYY s 5Y z
|
62 |
+
Yh 6Y xfYYY y qq i 2m Ydu Y6 Y vd sYY Yz cq vYYY eY 5Y0 b z q
|
63 |
+
YYz 6Y xywYYYY ZzqYt f YY Ydu Y6 Y vd sYY Yl 6d iYYY YYz 5YY Y z 9 g il
|
64 |
+
YYx YYo gfwYYYY ZxkY6 wf Z9YY Ydu xz6 Y6 Yi8 v4 sYY z YY 6d 4YYY uYYu o 5YY f Y zkyY 0 1Y z
|
65 |
+
YYw YY7 hfwYYYY 7 ZYkYY 1 z 7Y yYYYY YdY 8zY Y6 Y78 v4 sYYyt Y YY 6d 4YYYz3YYY o 4YYx e Yzx0xY Z 1Y 8
|
66 |
+
YYw YYY 5dwYYYYeYmZYmYY Y Y YY sYYYYY YdYl x 50Yo 9 Y6 YYY4 v5 vYY3tpY YY 6d 4YYYyYYYY m 3YYq Y Yzg0iY Y 4ZY 98 p
|
67 |
+
YYsjYYYYz 5Y7YYYYYYdYYYYY Y Y YY YYYYYY YcYY 5lY8YYa1 Y6 YYY2 v7YYYYYYpYz YYY 6d ZYYYtYYYY Yg 3YYo Y Yz8YZY xw Y YYY Y5 p Y
|
68 |
+
YY6YYYYYa YY9YYYYYYYYYYYYrYyYz YYe YYYYYY YYYYv YsYYYYYYY uY5 YYYY0 vYYYYYYYoYz YYY 6Y YYYYtYYYY Y7 3YYi rY wYz7YYYvkt YYYYY Y4yo4 1Yay
|
69 |
+
3YYdYYYYY8 YYYYYYYYYYYYYYYYYwYu YY3 YYYYYY YYYYn YeYYYYYYYZ uY5 YYYY0 uYYYYYYY6Yz YYY 6Y9 YYYYsYYYY Y58 3YYi6pY ZYz5YYY2e5 YYYYY Y2qeYz1YY9
|
70 |
+
YYY5YYYYY5 YYYYYYYYYYYYYYYYYuYu YY1 YYYYYY YYYY3 YYYYYYYYYY uYYYYYYY0 hYYYYYYY6YyrYYY 6YY YYYYYYYYY YYY 3YYfYsY YYz1YYY1Y2Y YYYYY YYq0YYYYYY
|
71 |
+
YYYYYYYYYY YYYYYYYYYYYYYYYYYsYuyYYYoYYYYYY1 YYYY2YYYYYYYYYYY vYYYYYYYY z4YYYYYYYYYYYYYYZg1YYYYYYYYYYYYzYYYz ZYYYYdYlYYYYYYYZYYY YYYYYzYYYYYYYYYY
|
72 |
+
YYYYYYYYYYYYY1YYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYZY112YYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY
|
73 |
+
YYY1YYYYYY YYYYYYYYYYYYYYYYYsYu YYYoYYYYYY4 YYYY2YYYYYYYYYYY vYYYYYYYY 7YYYYYYYZYYYYYYZv7YYYYYYYYYYYY YYYz 1YYYYdYmYYYYYYYZYYY YYYYYzYYYYYYYYYY
|
74 |
+
YYY6YYYYY6 YYYYYYYYYYYYYYYYYvYu YY1 YYYYYY YYYY3 YYYYYYYYYY uYYYYYYY0 hYYYYYYY6YyxYYY 6YY YYYYYYYYY YYY 3YYfYtY YYz3YYY1Y2Y YYYYY YYq0YYYYYY
|
75 |
+
cYYcYYYYY8 YYYYYYYYYYYYYYYYYxYu YY3 YYYYYY YYYYn YfYYYYYYYZ uY5 YYYY0 vYYYYYYY6Yz YYY 6Y9 YYYYsYYYY Y5n 3YYi8pY 1Yz6YYY3g7 YYYYY Y2qkYz1YYa
|
76 |
+
YY9YYYYYf YY9YYYYYYYYYYYYrYyYz YYi YYYYYY Y1YYv YsYZYYYYY uY5 YYYYb vYYYYYYYoYz YYY 6Y YYYYtYYYY Y7 3YYi rY wYz7YYYvku YYYYY Y3zo6 1Yn
|
77 |
+
YYskYYYYz 5Y8YYYYYYdYY2YY Y Y YY YYYYYY YcYY 5lY8YYd3 Y6 YYY4 v9YYYYYYpYz YYY 6d 1YYYtYYYY Yg 3YYp Y Yz8Y1Y yw Y YYY Y5 q Y
|
78 |
+
YYw YYY 5dwYYYYhYpZYmYY Y Y YY wYYYYY YdYr z 50Yo 0 Y6 YYY5 v5 vYY4tpY YY 6d 4YYYzYYYY m 3YYq Y Yzh0iY Y 4ZY 98 o
|
79 |
+
YYw YY7 hfwYYYY b ZYkYY 4 cY yYYYY YdY dzY Y6 Y78 v4 sYYzt Y YY 6d 4YYY 7YYY o 5YYx f Yzy0yY 3 1Y 8
|
80 |
+
YYx ZYo hfwYYYY ZykY7 wf Z9YY Ydu xz7 Y6 Ykb v3 sYY YY 6d 9YYY uYYy o 5YY f Y zpyY b 1Y z
|
81 |
+
YY 6Y xywYYYY ZzqYy f YY Ydu Y6 Y vd sYY Yo 6d jYYY 1Yz 5YY Y z 9 g lm
|
82 |
+
Yi 6Y ygYYY y qq i 9m Ydu Y6 Y vd sYY Y cs vYYY rY 5Yp e z t
|
83 |
+
6 YYY y s o Ys Yl xd sY Y r vYYY y 5Y z
|
84 |
+
6 YYY r Ys Yl xd sY Y eYY Y
|
85 |
+
e YYz r Ys Ym xd Y gYY
|
86 |
+
x 1Y z 6o ze 8 mYY
|
87 |
+
14 6o e o5
|
88 |
+
4 do t
|
89 |
+
4 dv v
|
90 |
+
h ex v
|
91 |
+
dx
|
92 |
+
ux
|
93 |
+
ux
|
94 |
+
x
|
95 |
+
x
|
96 |
+
x
|
97 |
+
y
|
98 |
+
|
99 |
+
|
100 |
+
```
|
101 |
+
|
102 |
+
## Models & Results
|
103 |
+
Mistral 7B 0.2 finetunes on ascii art. **As seen in the results, experiment didn't amount to much**
|
104 |
+
- 2 epochs - [0-hero/mistral-speech-to-text-preview](https://huggingface.co/0-hero/mistral-speech-to-text-preview/) - Loss pretty much flattened after this epoch
|
105 |
+
- ```
|
106 |
+
Metrics:
|
107 |
+
rouge-1: {'r': 0.12919024091165357, 'p': 0.1163312036605547, 'f': 0.11315199212991178}
|
108 |
+
rouge-2: {'r': 0.013705453572242508, 'p': 0.0137500428446463, 'f': 0.012676757505648992}
|
109 |
+
rouge-l: {'r': 0.11261286554140228, 'p': 0.09921920076529338, 'f': 0.09705621471622536}
|
110 |
+
length_correlation: 0.014470676120233311
|
111 |
+
avg_actual_length: 16.59
|
112 |
+
avg_pred_length: 21.46
|
113 |
+
exact_match_accuracy: 0.0
|
114 |
+
```
|
115 |
+
- 4 epochs - [0-hero/mistral-speech-to-text](https://huggingface.co/0-hero/mistral-speech-to-text/)
|
116 |
+
- ```
|
117 |
+
Metrics:
|
118 |
+
rouge-1: {'r': 0.11869828051815862, 'p': 0.11697319273190071, 'f': 0.11154343875398197}
|
119 |
+
rouge-2: {'r': 0.008572925612399297, 'p': 0.009040061245943597, 'f': 0.008369604666309954}
|
120 |
+
rouge-l: {'r': 0.10780857719316121, 'p': 0.10373665666448233, 'f': 0.09985384905943501}
|
121 |
+
length_correlation: -0.1500200314034927
|
122 |
+
avg_actual_length: 16.59
|
123 |
+
avg_pred_length: 18.32
|
124 |
+
exact_match_accuracy: 0.0
|
125 |
+
```
|
126 |
+
|
127 |
+
## Datasets
|
128 |
+
[lj_speech](https://huggingface.co/datasets/lj_speech) dataset used to convert audio waveforms into ASCII Art
|
129 |
+
- [0-hero/lj_speech_with_spectogram_conversations](https://huggingface.co/datasets/0-hero/lj_speech_with_spectogram_conversations) - ShareGPT style finetuning dataset with train, test split (this)
|
130 |
+
- [0-hero/lj_speech_with_spectogram](https://huggingface.co/datasets/0-hero/lj_speech_with_spectogram) - Raw dataset with ASCII Art
|
131 |
+
|
132 |
+
|