Instructions to use WindyWord/translate-lt-es with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use WindyWord/translate-lt-es 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="WindyWord/translate-lt-es")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("WindyWord/translate-lt-es", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Remove WindyScripture variants from current files (source-text licence review)
Browse filesRemoves herm0-scripture/ and scripture-ct2-int8/ from the current files; README and NOTICE updated; other variants unchanged. A precaution while the licences of the eBible source texts are reviewed, not a legal conclusion.
- NOTICE.md +2 -4
- README.md +2 -10
- herm0-scripture/config.json +0 -56
- herm0-scripture/generation_config.json +0 -16
- herm0-scripture/model.safetensors +0 -3
- herm0-scripture/source.spm +0 -3
- herm0-scripture/special_tokens_map.json +0 -23
- herm0-scripture/target.spm +0 -3
- herm0-scripture/tokenizer_config.json +0 -38
- herm0-scripture/training_meta.json +0 -15
- herm0-scripture/vocab.json +0 -0
- scripture-ct2-int8/config.json +0 -10
- scripture-ct2-int8/model.bin +0 -3
- scripture-ct2-int8/shared_vocabulary.json +0 -0
NOTICE.md
CHANGED
|
@@ -4,10 +4,8 @@
|
|
| 4 |
|
| 5 |
Weights derived from [Helsinki-NLP/opus-mt-lt-es](https://huggingface.co/Helsinki-NLP/opus-mt-lt-es) (OPUS-MT, Helsinki-NLP, University of Helsinki), licensed Apache-2.0. Windy variants are released under the same licence. The full licence text is in `LICENSE`.
|
| 6 |
|
| 7 |
-
##
|
| 8 |
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
This applies to the WindyScripture variants in this repository (`herm0-scripture/`, `scripture-ct2-int8/`). The other variants are derived from the upstream model under its licence. The `LICENSE` file covers those other variants only.
|
| 12 |
|
| 13 |
Published by Windstorm Labs. Catalogue entry: https://windytranslate.com/models/translate-lt-es
|
|
|
|
| 4 |
|
| 5 |
Weights derived from [Helsinki-NLP/opus-mt-lt-es](https://huggingface.co/Helsinki-NLP/opus-mt-lt-es) (OPUS-MT, Helsinki-NLP, University of Helsinki), licensed Apache-2.0. Windy variants are released under the same licence. The full licence text is in `LICENSE`.
|
| 6 |
|
| 7 |
+
## Removed variants
|
| 8 |
|
| 9 |
+
The WindyScripture variants (`herm0-scripture/`, `scripture-ct2-int8/`) were removed from this repository's current files on 2026-10-04 while the licences of their eBible source texts are reviewed. This is a precaution, not a legal conclusion. The `LICENSE` file covers the remaining variants.
|
|
|
|
|
|
|
| 10 |
|
| 11 |
Published by Windstorm Labs. Catalogue entry: https://windytranslate.com/models/translate-lt-es
|
README.md
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
---
|
| 2 |
-
license:
|
| 3 |
tags:
|
| 4 |
- translation
|
| 5 |
- marian
|
|
@@ -11,8 +11,6 @@ language:
|
|
| 11 |
- es
|
| 12 |
library_name: transformers
|
| 13 |
pipeline_tag: translation
|
| 14 |
-
license_name: under-review
|
| 15 |
-
license_link: https://huggingface.co/WindyWord/translate-lt-es/blob/main/NOTICE.md
|
| 16 |
base_model: Helsinki-NLP/opus-mt-lt-es
|
| 17 |
---
|
| 18 |
|
|
@@ -22,9 +20,7 @@ base_model: Helsinki-NLP/opus-mt-lt-es
|
|
| 22 |
>
|
| 23 |
> Canonical copy: https://huggingface.co/WindyTranslate/translate-lt-es
|
| 24 |
|
| 25 |
-
> **
|
| 26 |
-
>
|
| 27 |
-
> This applies to the WindyScripture variants in this repository (`herm0-scripture/`, `scripture-ct2-int8/`). The other variants are derived from the upstream model under its licence, shown in the Provenance & License section below.
|
| 28 |
|
| 29 |
**Translates Lithuanian → Spanish.**
|
| 30 |
|
|
@@ -40,8 +36,6 @@ Deployment formats in this repository (subfolders):
|
|
| 40 |
|---|---|
|
| 41 |
| `lora/` | **WindyStandard** — production baseline. Transformers format for GPU inference. |
|
| 42 |
| `lora-ct2-int8/` | **WindyStandard · CPU INT8** — CTranslate2 INT8 quantization of WindyStandard for CPU inference. |
|
| 43 |
-
| `herm0-scripture/` | **WindyScripture** — verse-aligned fine-tune on the eBible parallel corpus. Specialised for biblical text; not recommended for general translation. *Licence under review — see notice above.* |
|
| 44 |
-
| `scripture-ct2-int8/` | **WindyScripture · CPU INT8** — CTranslate2 INT8 quantization of WindyScripture. *Licence under review — see notice above.* |
|
| 45 |
|
| 46 |
### Quick usage
|
| 47 |
|
|
@@ -68,6 +62,4 @@ The [Windy Word](https://windyword.ai) apps are built on this model family.
|
|
| 68 |
|
| 69 |
Weights derived from [Helsinki-NLP/opus-mt-lt-es](https://huggingface.co/Helsinki-NLP/opus-mt-lt-es) (OPUS-MT, Helsinki-NLP, University of Helsinki), licensed Apache-2.0. Windy variants are released under the same licence.
|
| 70 |
|
| 71 |
-
Exception: the WindyScripture variants (`herm0-scripture/`, `scripture-ct2-int8/`) are under licence review — see the notice at the top of this card. Not for commercial use until that notice is removed.
|
| 72 |
-
|
| 73 |
Licence text: [`LICENSE`](LICENSE). Attribution: [`NOTICE.md`](NOTICE.md).
|
|
|
|
| 1 |
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
tags:
|
| 4 |
- translation
|
| 5 |
- marian
|
|
|
|
| 11 |
- es
|
| 12 |
library_name: transformers
|
| 13 |
pipeline_tag: translation
|
|
|
|
|
|
|
| 14 |
base_model: Helsinki-NLP/opus-mt-lt-es
|
| 15 |
---
|
| 16 |
|
|
|
|
| 20 |
>
|
| 21 |
> Canonical copy: https://huggingface.co/WindyTranslate/translate-lt-es
|
| 22 |
|
| 23 |
+
> **Note (2026-10-04):** the WindyScripture variants (`herm0-scripture/`, `scripture-ct2-int8/`) were removed from this repository's current files while the licences of their eBible source texts are reviewed. This is a precaution, not a legal conclusion. The other variants are unchanged.
|
|
|
|
|
|
|
| 24 |
|
| 25 |
**Translates Lithuanian → Spanish.**
|
| 26 |
|
|
|
|
| 36 |
|---|---|
|
| 37 |
| `lora/` | **WindyStandard** — production baseline. Transformers format for GPU inference. |
|
| 38 |
| `lora-ct2-int8/` | **WindyStandard · CPU INT8** — CTranslate2 INT8 quantization of WindyStandard for CPU inference. |
|
|
|
|
|
|
|
| 39 |
|
| 40 |
### Quick usage
|
| 41 |
|
|
|
|
| 62 |
|
| 63 |
Weights derived from [Helsinki-NLP/opus-mt-lt-es](https://huggingface.co/Helsinki-NLP/opus-mt-lt-es) (OPUS-MT, Helsinki-NLP, University of Helsinki), licensed Apache-2.0. Windy variants are released under the same licence.
|
| 64 |
|
|
|
|
|
|
|
| 65 |
Licence text: [`LICENSE`](LICENSE). Attribution: [`NOTICE.md`](NOTICE.md).
|
herm0-scripture/config.json
DELETED
|
@@ -1,56 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"_name_or_path": "/home/user1-gpu/Desktop/grants_folder/windy-pro/models/windy-pair-lt-es/base",
|
| 3 |
-
"activation_dropout": 0.0,
|
| 4 |
-
"activation_function": "swish",
|
| 5 |
-
"add_bias_logits": false,
|
| 6 |
-
"add_final_layer_norm": false,
|
| 7 |
-
"architectures": [
|
| 8 |
-
"MarianMTModel"
|
| 9 |
-
],
|
| 10 |
-
"attention_dropout": 0.0,
|
| 11 |
-
"bos_token_id": 0,
|
| 12 |
-
"classif_dropout": 0.0,
|
| 13 |
-
"classifier_dropout": 0.0,
|
| 14 |
-
"d_model": 512,
|
| 15 |
-
"decoder_attention_heads": 8,
|
| 16 |
-
"decoder_ffn_dim": 2048,
|
| 17 |
-
"decoder_layerdrop": 0.0,
|
| 18 |
-
"decoder_layers": 6,
|
| 19 |
-
"decoder_start_token_id": 59510,
|
| 20 |
-
"decoder_vocab_size": 59511,
|
| 21 |
-
"dropout": 0.1,
|
| 22 |
-
"encoder_attention_heads": 8,
|
| 23 |
-
"encoder_ffn_dim": 2048,
|
| 24 |
-
"encoder_layerdrop": 0.0,
|
| 25 |
-
"encoder_layers": 6,
|
| 26 |
-
"eos_token_id": 0,
|
| 27 |
-
"extra_pos_embeddings": 59511,
|
| 28 |
-
"forced_eos_token_id": 0,
|
| 29 |
-
"id2label": {
|
| 30 |
-
"0": "LABEL_0",
|
| 31 |
-
"1": "LABEL_1",
|
| 32 |
-
"2": "LABEL_2"
|
| 33 |
-
},
|
| 34 |
-
"init_std": 0.02,
|
| 35 |
-
"is_encoder_decoder": true,
|
| 36 |
-
"label2id": {
|
| 37 |
-
"LABEL_0": 0,
|
| 38 |
-
"LABEL_1": 1,
|
| 39 |
-
"LABEL_2": 2
|
| 40 |
-
},
|
| 41 |
-
"max_length": null,
|
| 42 |
-
"max_position_embeddings": 512,
|
| 43 |
-
"model_type": "marian",
|
| 44 |
-
"normalize_before": false,
|
| 45 |
-
"normalize_embedding": false,
|
| 46 |
-
"num_beams": null,
|
| 47 |
-
"num_hidden_layers": 6,
|
| 48 |
-
"pad_token_id": 59510,
|
| 49 |
-
"scale_embedding": true,
|
| 50 |
-
"share_encoder_decoder_embeddings": true,
|
| 51 |
-
"static_position_embeddings": true,
|
| 52 |
-
"torch_dtype": "float32",
|
| 53 |
-
"transformers_version": "4.45.1",
|
| 54 |
-
"use_cache": true,
|
| 55 |
-
"vocab_size": 59511
|
| 56 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/generation_config.json
DELETED
|
@@ -1,16 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"bad_words_ids": [
|
| 3 |
-
[
|
| 4 |
-
59510
|
| 5 |
-
]
|
| 6 |
-
],
|
| 7 |
-
"bos_token_id": 0,
|
| 8 |
-
"decoder_start_token_id": 59510,
|
| 9 |
-
"eos_token_id": 0,
|
| 10 |
-
"forced_eos_token_id": 0,
|
| 11 |
-
"max_length": 512,
|
| 12 |
-
"num_beams": 6,
|
| 13 |
-
"pad_token_id": 59510,
|
| 14 |
-
"renormalize_logits": true,
|
| 15 |
-
"transformers_version": "4.45.1"
|
| 16 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/model.safetensors
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:ddc4113f8671acb393c6b25ff900155e0096a250bb73f210ef913add6d1aff43
|
| 3 |
-
size 298699612
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/source.spm
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:c6bd310045566d83eab4047498c7310dec0be8407b582bd35de0397eba9685ed
|
| 3 |
-
size 838643
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/special_tokens_map.json
DELETED
|
@@ -1,23 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"eos_token": {
|
| 3 |
-
"content": "</s>",
|
| 4 |
-
"lstrip": false,
|
| 5 |
-
"normalized": false,
|
| 6 |
-
"rstrip": false,
|
| 7 |
-
"single_word": false
|
| 8 |
-
},
|
| 9 |
-
"pad_token": {
|
| 10 |
-
"content": "<pad>",
|
| 11 |
-
"lstrip": false,
|
| 12 |
-
"normalized": false,
|
| 13 |
-
"rstrip": false,
|
| 14 |
-
"single_word": false
|
| 15 |
-
},
|
| 16 |
-
"unk_token": {
|
| 17 |
-
"content": "<unk>",
|
| 18 |
-
"lstrip": false,
|
| 19 |
-
"normalized": false,
|
| 20 |
-
"rstrip": false,
|
| 21 |
-
"single_word": false
|
| 22 |
-
}
|
| 23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/target.spm
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:a2eeb42fe530d4ffeb0b0d2a5d71e2d9bf04f6156af8b4466dbefb98264093a7
|
| 3 |
-
size 830857
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/tokenizer_config.json
DELETED
|
@@ -1,38 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"added_tokens_decoder": {
|
| 3 |
-
"0": {
|
| 4 |
-
"content": "</s>",
|
| 5 |
-
"lstrip": false,
|
| 6 |
-
"normalized": false,
|
| 7 |
-
"rstrip": false,
|
| 8 |
-
"single_word": false,
|
| 9 |
-
"special": true
|
| 10 |
-
},
|
| 11 |
-
"1": {
|
| 12 |
-
"content": "<unk>",
|
| 13 |
-
"lstrip": false,
|
| 14 |
-
"normalized": false,
|
| 15 |
-
"rstrip": false,
|
| 16 |
-
"single_word": false,
|
| 17 |
-
"special": true
|
| 18 |
-
},
|
| 19 |
-
"59510": {
|
| 20 |
-
"content": "<pad>",
|
| 21 |
-
"lstrip": false,
|
| 22 |
-
"normalized": false,
|
| 23 |
-
"rstrip": false,
|
| 24 |
-
"single_word": false,
|
| 25 |
-
"special": true
|
| 26 |
-
}
|
| 27 |
-
},
|
| 28 |
-
"clean_up_tokenization_spaces": false,
|
| 29 |
-
"eos_token": "</s>",
|
| 30 |
-
"model_max_length": 512,
|
| 31 |
-
"pad_token": "<pad>",
|
| 32 |
-
"separate_vocabs": false,
|
| 33 |
-
"source_lang": "lit",
|
| 34 |
-
"sp_model_kwargs": {},
|
| 35 |
-
"target_lang": "spa",
|
| 36 |
-
"tokenizer_class": "MarianTokenizer",
|
| 37 |
-
"unk_token": "<unk>"
|
| 38 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/training_meta.json
DELETED
|
@@ -1,15 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"model_id": "windy-pair-lt-es",
|
| 3 |
-
"doctor": "herm0",
|
| 4 |
-
"procedure": "ebible_finetune_v3",
|
| 5 |
-
"timestamp": "2026-03-28T04:06:25.955379+00:00",
|
| 6 |
-
"data_source": "ebible.org",
|
| 7 |
-
"training_pairs": 7103,
|
| 8 |
-
"base_loss": 9.9188032913208,
|
| 9 |
-
"ft_loss": 2.3760746160149573,
|
| 10 |
-
"improvement_pct": 76.04474505413287,
|
| 11 |
-
"epochs": 1,
|
| 12 |
-
"learning_rate": 2e-05,
|
| 13 |
-
"batch_size": 16,
|
| 14 |
-
"max_length": 128
|
| 15 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
herm0-scripture/vocab.json
DELETED
|
The diff for this file is too large to render.
See raw diff
|
|
|
scripture-ct2-int8/config.json
DELETED
|
@@ -1,10 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"add_source_bos": false,
|
| 3 |
-
"add_source_eos": false,
|
| 4 |
-
"bos_token": "<s>",
|
| 5 |
-
"decoder_start_token": "</s>",
|
| 6 |
-
"eos_token": "</s>",
|
| 7 |
-
"layer_norm_epsilon": null,
|
| 8 |
-
"multi_query_attention": false,
|
| 9 |
-
"unk_token": "<unk>"
|
| 10 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
scripture-ct2-int8/model.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:e140e03b31a256a11dcd44b5a6311ad1048e948268ee2b0283ab54fb6a370da2
|
| 3 |
-
size 76712835
|
|
|
|
|
|
|
|
|
|
|
|
scripture-ct2-int8/shared_vocabulary.json
DELETED
|
The diff for this file is too large to render.
See raw diff
|
|
|