I帽igo L贸pez-Riob贸o Botana
commited on
Commit
路
cf9337c
1
Parent(s):
e6bff60
Update README.md
Browse files
README.md
CHANGED
@@ -6,14 +6,17 @@ pipeline_tag: text-generation
|
|
6 |
---
|
7 |
|
8 |
# DialoGPT-medium-spanish-chitchat
|
|
|
9 |
## Description
|
10 |
|
11 |
-
This is a **transformer-decoder** [gpt2 model](https://huggingface.co/gpt2), adapted for **single turn dialogue tasks**. We fine-tuned a [DialoGPT-medium](https://huggingface.co/microsoft/DialoGPT-medium) model from Microsoft, following the CLM (Causal Language Modelling) objective.
|
12 |
We used one of the datasets available in the [Bot Framework Tools repository](https://github.com/microsoft/botframework-cli). We processed [the professional-styled personality chat dataset in Spanish](https://github.com/microsoft/botframework-cli/blob/main/packages/qnamaker/docs/chit-chat-dataset.md), the file is available [here](https://qnamakerstore.blob.core.windows.net/qnamakerdata/editorial/spanish/qna_chitchat_professional.tsv)
|
|
|
13 |
---
|
|
|
14 |
## Example inference script
|
15 |
|
16 |
-
Check at this example script to run this model in inference mode
|
17 |
|
18 |
```python
|
19 |
import torch
|
@@ -39,9 +42,12 @@ for i in range(CHAT_TURNS):
|
|
39 |
print(f"Step - {i} >> DialoGPT-spanish model answer -> {step_model_answer}")
|
40 |
|
41 |
```
|
|
|
42 |
---
|
|
|
43 |
## Examples
|
44 |
-
|
|
|
45 |
|
46 |
|Role | Response |
|
47 |
|---------|--------|
|
@@ -59,6 +65,16 @@ for i in range(CHAT_TURNS):
|
|
59 |
| Bot | En realidad no tengo tanta gracia. |
|
60 |
|User | Jajajajajajajaja |
|
61 |
| Bot | 隆Me alegro de que te r铆as! |
|
|
|
|
|
|
|
62 |
## Fine-tuning in different dataset
|
63 |
|
64 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
---
|
7 |
|
8 |
# DialoGPT-medium-spanish-chitchat
|
9 |
+
|
10 |
## Description
|
11 |
|
12 |
+
This is a **transformer-decoder** [gpt2 model](https://huggingface.co/gpt2), adapted for **single turn dialogue tasks**. We fine-tuned a 345M parameters [DialoGPT-medium](https://huggingface.co/microsoft/DialoGPT-medium) model from Microsoft, following the CLM (Causal Language Modelling) objective.
|
13 |
We used one of the datasets available in the [Bot Framework Tools repository](https://github.com/microsoft/botframework-cli). We processed [the professional-styled personality chat dataset in Spanish](https://github.com/microsoft/botframework-cli/blob/main/packages/qnamaker/docs/chit-chat-dataset.md), the file is available [here](https://qnamakerstore.blob.core.windows.net/qnamakerdata/editorial/spanish/qna_chitchat_professional.tsv)
|
14 |
+
|
15 |
---
|
16 |
+
|
17 |
## Example inference script
|
18 |
|
19 |
+
### Check at this example script to run this model in inference mode
|
20 |
|
21 |
```python
|
22 |
import torch
|
|
|
42 |
print(f"Step - {i} >> DialoGPT-spanish model answer -> {step_model_answer}")
|
43 |
|
44 |
```
|
45 |
+
|
46 |
---
|
47 |
+
|
48 |
## Examples
|
49 |
+
|
50 |
+
### Single-turn generation examples from an interactive environment
|
51 |
|
52 |
|Role | Response |
|
53 |
|---------|--------|
|
|
|
65 |
| Bot | En realidad no tengo tanta gracia. |
|
66 |
|User | Jajajajajajajaja |
|
67 |
| Bot | 隆Me alegro de que te r铆as! |
|
68 |
+
|
69 |
+
---
|
70 |
+
|
71 |
## Fine-tuning in different dataset
|
72 |
|
73 |
+
If you want to fine-tune this model, we recommend you to start from the [DialoGPT model](https://huggingface.co/microsoft/DialoGPT-medium).
|
74 |
+
You can check the [original GitHub repository](https://github.com/microsoft/DialoGPT).
|
75 |
+
|
76 |
+
## Limitations
|
77 |
+
|
78 |
+
- This model is intended to be used **just for single-turn chitchat conversations in Spanish**
|
79 |
+
- This model's generation capabilities are limited to the extent of the aforementioned fine-tuning dataset
|
80 |
+
- This model generates short answers, providing general context dialogue in a professional style
|