Promote lora build from WindstormLabs/translate-lua-es, with attribution and change statement

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+source.spm filter=lfs diff=lfs merge=lfs -text
+target.spm filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,57 @@
+---
+license: apache-2.0
+base_model: Helsinki-NLP/opus-mt-lua-es
+library_name: transformers
+pipeline_tag: translation
+tags:
+- translation
+- marian
+---
+
+# translate-lua-es
+
+Machine translation model, `lua` to `es`, published by Windstorm Labs.
+
+## Attribution
+
+Derived from [`Helsinki-NLP/opus-mt-lua-es`](https://huggingface.co/Helsinki-NLP/opus-mt-lua-es), licensed **apache-2.0**.
+Windstorm Labs did not train the original model. This notice provides the attribution the
+licence requires.
+
+## What was changed
+
+**The weights in this repository have been modified from the original.** A LoRA fine-tune trained on parallel corpus data and merged into the base weights.
+
+| | |
+|---|---|
+| Method | `lora-fog` |
+| Tensors modified | **36 of 254** |
+| Max absolute weight delta | **6.614e-05** |
+| Verified | tensor-by-tensor against the upstream original, 2026-07-26 |
+
+The comparison is tensor-level rather than file-level: safetensors and PyTorch `.bin`
+containers hash differently even when the tensors inside are identical, so a file-hash
+mismatch would prove nothing.
+
+## Contents
+
+A single transformers build in safetensors format, loadable directly:
+
+```python
+from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+
+tok = AutoTokenizer.from_pretrained("WindyTranslate/translate-lua-es")
+model = AutoModelForSeq2SeqLM.from_pretrained("WindyTranslate/translate-lua-es")
+
+batch = tok(["<your text here>"], return_tensors="pt", padding=True)
+print(tok.batch_decode(model.generate(**batch), skip_special_tokens=True))
+```
+
+Other builds of this pair, including CTranslate2 INT8, are not included in this repository.
+
+## Limitations
+
+- A single language direction: `lua` to `es`.
+- **No quality benchmark has been run on this pair.** The weights are verified to differ from
+  the original; that is a statement about provenance, not about translation quality.
+- Quality is not monotonic with model size or with the amount of fine-tuning applied.

config.json

@@ -0,0 +1,56 @@
+{
+  "_name_or_path": "Helsinki-NLP/opus-mt-lua-es",
+  "_num_labels": 3,
+  "activation_dropout": 0.0,
+  "activation_function": "swish",
+  "add_bias_logits": false,
+  "add_final_layer_norm": false,
+  "architectures": [
+    "MarianMTModel"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "classif_dropout": 0.0,
+  "classifier_dropout": 0.0,
+  "d_model": 512,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 2048,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "decoder_start_token_id": 60812,
+  "decoder_vocab_size": 60813,
+  "dropout": 0.1,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_token_id": 0,
+  "forced_eos_token_id": 0,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1",
+    "2": "LABEL_2"
+  },
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "LABEL_0": 0,
+    "LABEL_1": 1,
+    "LABEL_2": 2
+  },
+  "max_length": null,
+  "max_position_embeddings": 512,
+  "model_type": "marian",
+  "normalize_before": false,
+  "normalize_embedding": false,
+  "num_beams": null,
+  "num_hidden_layers": 6,
+  "pad_token_id": 60812,
+  "scale_embedding": true,
+  "share_encoder_decoder_embeddings": true,
+  "static_position_embeddings": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.45.1",
+  "use_cache": true,
+  "vocab_size": 60813
+}

generation_config.json

@@ -0,0 +1,16 @@
+{
+  "bad_words_ids": [
+    [
+      60812
+    ]
+  ],
+  "bos_token_id": 0,
+  "decoder_start_token_id": 60812,
+  "eos_token_id": 0,
+  "forced_eos_token_id": 0,
+  "max_length": 512,
+  "num_beams": 6,
+  "pad_token_id": 60812,
+  "renormalize_logits": true,
+  "transformers_version": "4.45.1"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6fe96f122059fb21b88f14880e8dda86761b644eb85ec577bb5a069be981a909
+size 301371316

source.spm

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbe7059341233a855bd72e95b2fd84acb711b6acd48f5e46b604af5edbaff378
+size 823233

special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

target.spm

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5bf28c020a04fc922f671f5e121486cb848924b250b7ab78803dc5dd64f9a3b
+size 845457

tokenizer_config.json

@@ -0,0 +1,38 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "60812": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "separate_vocabs": false,
+  "source_lang": "lua",
+  "sp_model_kwargs": {},
+  "target_lang": "es",
+  "tokenizer_class": "MarianTokenizer",
+  "unk_token": "<unk>"
+}