Instructions to use anocha-mcomp/lora_model with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use anocha-mcomp/lora_model with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("anocha-mcomp/lora_model", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Unsloth Desktop
Update model card from Confluence: Qwen-SEA-LION-v4-32B-IT
Browse files
qwen-sea-lion-v4-32b-it.md
CHANGED
|
@@ -16,7 +16,7 @@ Qwen-SEA-LION-v4-32B-IT inherits the following features from Qwen3-32B:
|
|
| 16 |
|
| 17 |
SEA-LION stands for *Southeast Asian Languages In One Network*.
|
| 18 |
|
| 19 |
-
We performed continued pre-training in English and SEA languages on Qwen3-32B, a decoder model using the
|
| 20 |
|
| 21 |
For tokenization, the model employs the default tokenizer used in Qwen3-32B.
|
| 22 |
|
|
@@ -27,7 +27,7 @@ For tokenization, the model employs the default tokenizer used in Qwen3-32B.
|
|
| 27 |
- **Context Length:** 128k tokens
|
| 28 |
- **Language(s) (NLP):** Burmese, English, Indonesian, Khmer, Lao, Malay, Mandarin, Tagalog, Tamil, Thai, and Vietnamese
|
| 29 |
- **License:** [Qwen Terms of Service](https://qwen.ai/termsservice) / [Qwen Usage Policy](https://qwen.ai/usagepolicy)
|
| 30 |
-
- **Continue pretrained from model:** [
|
| 31 |
|
| 32 |
## Uses
|
| 33 |
|
|
@@ -56,7 +56,7 @@ Use the code below to get started with the model using the 🤗 Transformers lib
|
|
| 56 |
```
|
| 57 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 58 |
|
| 59 |
-
model_name = "aisingapore/Qwen-SEA-LION-v4-32B"
|
| 60 |
|
| 61 |
# load the tokenizer and the model
|
| 62 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
|
@@ -104,13 +104,13 @@ print("content:", content)
|
|
| 104 |
|
| 105 |
**Training Datasets**
|
| 106 |
|
| 107 |
-
The instruction fine-tuning dataset combines
|
| 108 |
|
| 109 |
#### Training Procedure
|
| 110 |
|
| 111 |
**Training regime**
|
| 112 |
|
| 113 |
-
Our post-training workflow consists of
|
| 114 |
|
| 115 |
## Uses
|
| 116 |
|
|
@@ -130,7 +130,7 @@ Our post-training workflow consists of multiple stages: instruction fine-tuning,
|
|
| 130 |
|
| 131 |
*Additional Remarks:*
|
| 132 |
|
| 133 |
-
- TTFT and
|
| 134 |
- Reported results are the median (p50) values, calculated across 10 requests.
|
| 135 |
(11 requests were run and the first result was dropped, to eliminate cold-start delays)
|
| 136 |
- Model size taken from vLLM upon loading
|
|
@@ -192,7 +192,7 @@ The following metrics were used:
|
|
| 192 |
|
| 193 |
### Results
|
| 194 |
|
| 195 |
-
[Attachment: image-20251017-
|
| 196 |
|
| 197 |
For details on Qwen-SEA-LION-v4-27B-IT performance, please refer to the SEA-HELM leaderboard, <https://leaderboard.sea-lion.ai/> .
|
| 198 |
|
|
@@ -204,4 +204,4 @@ For more info, please contact us using this [sealion@aisingapore.org](mailto:sea
|
|
| 204 |
|
| 205 |
## Team
|
| 206 |
|
| 207 |
-
Ahn Jeongmi, Antonyrex Sajeban, Chan Hok Teng Adwin, Cheng Zi Yi Nicholas, Choa Hsueh Mei Esther, Heng Jonathan, Huang Yuli, Hulagadri Adithya Venkatadri, Jann Railey Estrada Montalan, Kang Siow Wei Bryan, Lau Wayne, Lee Chwan Ren, Leong Wai Yi, Leong Wei Qi, Limkonchotiwat Peerat, Muhammad Ridzuan Bin Mokhtar, Nagarajan Karthik, Ng Boon Cheong Raymond, Ngee Chia Tai, Ngui Jian Gang, Nguyen Thanh Ngan, Ong Jin Jie Brandon, Ong Tat-Wee David, Ong Zhi Hao, Pereira Mark, Rengarajan Hamsawardhini, Susanto Yosephine, Sutaveephamochanon Anocha, Tan Choon Meng, Tan Chor Phin Evelyn, Tan Siao Wei Jessica, Tan Yixian, Tee Jun Yun, Teng Kok Wai Walter, Teo Eng Sipp Leslie, Tjhi William, Yeo Yeow Tong, Yong Xianbin, Zhang Zhou, Liew Rachel, Liu Bing Jie Darius,
|
|
|
|
| 16 |
|
| 17 |
SEA-LION stands for *Southeast Asian Languages In One Network*.
|
| 18 |
|
| 19 |
+
We performed continued pre-training in English and SEA languages on Qwen3-32B, a decoder model using the Qwen3 architecture, and post-training to create Qwen-SEA-LION-v4-32B-IT.
|
| 20 |
|
| 21 |
For tokenization, the model employs the default tokenizer used in Qwen3-32B.
|
| 22 |
|
|
|
|
| 27 |
- **Context Length:** 128k tokens
|
| 28 |
- **Language(s) (NLP):** Burmese, English, Indonesian, Khmer, Lao, Malay, Mandarin, Tagalog, Tamil, Thai, and Vietnamese
|
| 29 |
- **License:** [Qwen Terms of Service](https://qwen.ai/termsservice) / [Qwen Usage Policy](https://qwen.ai/usagepolicy)
|
| 30 |
+
- **Continue pretrained from model:** [Qwen3-32B](https://huggingface.co/Qwen/Qwen3-32B)
|
| 31 |
|
| 32 |
## Uses
|
| 33 |
|
|
|
|
| 56 |
```
|
| 57 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 58 |
|
| 59 |
+
model_name = "aisingapore/Qwen-SEA-LION-v4-32B-IT"
|
| 60 |
|
| 61 |
# load the tokenizer and the model
|
| 62 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
|
|
|
| 104 |
|
| 105 |
**Training Datasets**
|
| 106 |
|
| 107 |
+
The instruction fine-tuning dataset combines a collection of OSS & synthetic data.
|
| 108 |
|
| 109 |
#### Training Procedure
|
| 110 |
|
| 111 |
**Training regime**
|
| 112 |
|
| 113 |
+
Our post-training workflow consists of instruction fine-tuning and model merging.
|
| 114 |
|
| 115 |
## Uses
|
| 116 |
|
|
|
|
| 130 |
|
| 131 |
*Additional Remarks:*
|
| 132 |
|
| 133 |
+
- TTFT and Tokens per Second: measured with vLLM on localhost and concurrency = 1.
|
| 134 |
- Reported results are the median (p50) values, calculated across 10 requests.
|
| 135 |
(11 requests were run and the first result was dropped, to eliminate cold-start delays)
|
| 136 |
- Model size taken from vLLM upon loading
|
|
|
|
| 192 |
|
| 193 |
### Results
|
| 194 |
|
| 195 |
+
[Attachment: image-20251017-071644.png]Leaderboard snapshotted on 17 Oct 2025 10:44am
|
| 196 |
|
| 197 |
For details on Qwen-SEA-LION-v4-27B-IT performance, please refer to the SEA-HELM leaderboard, <https://leaderboard.sea-lion.ai/> .
|
| 198 |
|
|
|
|
| 204 |
|
| 205 |
## Team
|
| 206 |
|
| 207 |
+
Ahn Jeongmi, Antonyrex Sajeban, Chan Hok Teng Adwin, Cheng Zi Yi Nicholas, Choa Hsueh Mei Esther, Heng Jonathan, Huang Yuli, Hulagadri Adithya Venkatadri, Jann Railey Estrada Montalan, Kang Siow Wei Bryan, Lau Wayne, Lee Chwan Ren, Leong Wai Yi, Leong Wei Qi, Limkonchotiwat Peerat, Muhammad Ridzuan Bin Mokhtar, Nagarajan Karthik, Ng Boon Cheong Raymond, Ngee Chia Tai, Ngui Jian Gang, Nguyen Thanh Ngan, Ong Jin Jie Brandon, Ong Tat-Wee David, Ong Zhi Hao, Pereira Mark, Rengarajan Hamsawardhini, Susanto Yosephine, Sutaveephamochanon Anocha, Tan Choon Meng, Tan Chor Phin Evelyn, Tan Siao Wei Jessica, Tan Yixian, Tee Jun Yun, Teng Kok Wai Walter, Teo Eng Sipp Leslie, Tjhi William, Yeo Yeow Tong, Yong Xianbin, Zhang Zhou, Liew Rachel, Liu Bing Jie Darius, Tep Kilian Rithi (GoTo)
|