Translation
Transformers
Safetensors
Burmese
English
myanmartranslator
feature-extraction
Translation
Safetensors
Burmese
English
MyanmarLLM
flagship
custom_code
myanmar
Instructions to use freococo/MyanmarTranslator-210M-Pro with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use freococo/MyanmarTranslator-210M-Pro with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # pip install "transformers<5.0.0" from transformers import pipeline pipe = pipeline("translation", model="freococo/MyanmarTranslator-210M-Pro", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("freococo/MyanmarTranslator-210M-Pro", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -12,6 +12,10 @@ tags:
|
|
| 12 |
- MyanmarLLM
|
| 13 |
- flagship
|
| 14 |
- custom_code
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
---
|
| 16 |
|
| 17 |
# MyanmarTranslator-210M-Pro
|
|
@@ -19,4 +23,4 @@ tags:
|
|
| 19 |
Flagship `210.90M`-parameter **8-Layer Bidirectional Encoder + 8-Layer Cross-Attentive Decoder** (`d_model=768`, `n_heads=12`, `ffn_dim=2816` SwiGLU) for **English ↔ Burmese (`my`)**, warm-started via All-Layer Net2Wider+Net2Deeper expansion from `MyanmarTranslator-84M-Base`.
|
| 20 |
|
| 21 |
- **Validation NLL:** `1.6724` (**`Val PPL: 5.33`**)
|
| 22 |
-
- **Reproducibility Artifacts:** `training_codes.txt` & `trainingcelloutputwitheval.txt`
|
|
|
|
| 12 |
- MyanmarLLM
|
| 13 |
- flagship
|
| 14 |
- custom_code
|
| 15 |
+
- myanmar
|
| 16 |
+
datasets:
|
| 17 |
+
- HuggingFaceFW/finetranslations
|
| 18 |
+
pipeline_tag: translation
|
| 19 |
---
|
| 20 |
|
| 21 |
# MyanmarTranslator-210M-Pro
|
|
|
|
| 23 |
Flagship `210.90M`-parameter **8-Layer Bidirectional Encoder + 8-Layer Cross-Attentive Decoder** (`d_model=768`, `n_heads=12`, `ffn_dim=2816` SwiGLU) for **English ↔ Burmese (`my`)**, warm-started via All-Layer Net2Wider+Net2Deeper expansion from `MyanmarTranslator-84M-Base`.
|
| 24 |
|
| 25 |
- **Validation NLL:** `1.6724` (**`Val PPL: 5.33`**)
|
| 26 |
+
- **Reproducibility Artifacts:** `training_codes.txt` & `trainingcelloutputwitheval.txt`
|