Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,60 @@
|
|
1 |
-
---
|
2 |
-
license: bigscience-bloom-rail-1.0
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: bigscience-bloom-rail-1.0
|
3 |
+
datasets:
|
4 |
+
- swap-uniba/itwiki-march-2024
|
5 |
+
language:
|
6 |
+
- it
|
7 |
+
tags:
|
8 |
+
- bloom
|
9 |
+
- italian
|
10 |
+
---
|
11 |
+
|
12 |
+
# Model Card for Model ID
|
13 |
+
|
14 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
15 |
+
|
16 |
+
The model is obtained by performing language adaptation on the original bloom-1b7 model.
|
17 |
+
In detail, we continued the pre-training on Italian-specific data without adaptation of the vocabulary.
|
18 |
+
We use about 2.8M documents obtained from Italian Wikimedia dumps (swap-uniba/itwiki-march-2024). The model is trained for one epoch using LoRA and SFT.
|
19 |
+
|
20 |
+
## Model Details
|
21 |
+
|
22 |
+
### Model Description
|
23 |
+
|
24 |
+
<!-- Provide a longer summary of what this model is. -->
|
25 |
+
|
26 |
+
- **Developed by:** SWAP Research Group, Department of Computer Science, University of Bari Aldo Moro
|
27 |
+
- **Model type:** BLOOM
|
28 |
+
- **Language(s) (NLP):** Italian
|
29 |
+
- **License:** bigscience-bloom-rail-1.0
|
30 |
+
- **Finetuned from model [optional]:** bloom-1b7
|
31 |
+
|
32 |
+
## Training Details
|
33 |
+
|
34 |
+
### Training Data
|
35 |
+
|
36 |
+
2.8M documents obtained from Italian Wikimedia dumps (swap-uniba/itwiki-march-2024).
|
37 |
+
|
38 |
+
### Training Procedure
|
39 |
+
|
40 |
+
LoRA and SFT.
|
41 |
+
|
42 |
+
#### Training Hyperparameters
|
43 |
+
|
44 |
+
- **Training regime:** fp16
|
45 |
+
|
46 |
+
## Citation [optional]
|
47 |
+
|
48 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
49 |
+
|
50 |
+
**BibTeX:**
|
51 |
+
|
52 |
+
**APA:**
|
53 |
+
|
54 |
+
## Model Card Authors [optional]
|
55 |
+
|
56 |
+
Pierpaolo Basile, University of Bari Aldo Moro, Italy.
|
57 |
+
|
58 |
+
## Model Card Contact
|
59 |
+
|
60 |
+
Pierpaolo Basile, University of Bari Aldo Moro, Italy.
|