sneakyfree commited on
Commit
f6cfff5
·
verified ·
1 Parent(s): f3f1466

clone from WindyWord/translate-aed-es via local upload (ADR-039 Phase C)

Browse files
Files changed (1) hide show
  1. README.md +74 -0
README.md ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ tags:
4
+ - translation
5
+ - marian
6
+ - windyword
7
+ - argentine-sign-language
8
+ - spanish
9
+ language:
10
+ - aed
11
+ - es
12
+ library_name: transformers
13
+ pipeline_tag: translation
14
+ ---
15
+
16
+ # WindyWord.ai Translation — Argentine Sign Language → Spanish
17
+
18
+ **Translates Argentine Sign Language → Spanish.**
19
+
20
+ **Quality Rating: ⭐⭐½ (2.5★ Basic)**
21
+
22
+ Part of the [WindyWord.ai](https://windyword.ai) translation fleet — 1,800+ proprietary language pairs.
23
+
24
+ ## Quality & Pricing Tier
25
+
26
+ - **5-star rating:** 2.5★ ⭐⭐½
27
+ - **Tier:** Basic
28
+ - **Composite score:** 52.6 / 100
29
+ - **Rated via:** Grand Rounds v2 — an 8-test stress battery (paragraphs, multi-paragraph, native input, domain stress, edge cases, round-trip fidelity, speed, and consistency checks)
30
+
31
+ ## Available Variants
32
+
33
+ This repository contains multiple deployment formats. Pick the one that matches your use case:
34
+
35
+ | Variant | Description |
36
+ |---|---|
37
+ | `lora/` | **WindyStandard** — our proprietary production baseline. Stable, reliable, optimized for GPU inference. |
38
+ | `lora-ct2-int8/` | **WindyStandard · CPU INT8** — CTranslate2 quantized version of WindyStandard. ~25% of the size, 2–4× faster on CPU, no measurable quality loss. |
39
+
40
+ ### Quick usage
41
+
42
+ **Transformers (PyTorch):**
43
+ ```python
44
+ from transformers import MarianMTModel, MarianTokenizer
45
+ tokenizer = MarianTokenizer.from_pretrained("WindyWord/translate-aed-es", subfolder="lora")
46
+ model = MarianMTModel.from_pretrained("WindyWord/translate-aed-es", subfolder="lora")
47
+ ```
48
+
49
+ **CTranslate2 (fast CPU inference):**
50
+ ```python
51
+ import ctranslate2
52
+ translator = ctranslate2.Translator("path/to/translate-aed-es/lora-ct2-int8")
53
+ ```
54
+
55
+ ## Commercial Use
56
+
57
+ The WindyWord.ai platform provides:
58
+ - **Mobile apps** (iOS, Android — coming soon)
59
+ - **Real-time voice-to-text-to-translation** pipeline
60
+ - **API access** with premium model quality
61
+ - **Offline deployment** support
62
+
63
+ Visit [windyword.ai](https://windyword.ai) for apps and commercial API access.
64
+
65
+ ---
66
+
67
+ ## Provenance & License
68
+
69
+ Weights derived from the OPUS-MT project ([Helsinki-NLP/opus-mt-aed-es](https://huggingface.co/Helsinki-NLP/opus-mt-aed-es)) under CC-BY-4.0. WindyStandard, WindyEnhanced, and WindyScripture variants are proprietary to WindyWord.ai, independently trained and quality-certified via our Grand Rounds v2 test battery.
70
+
71
+ Licensed CC-BY-4.0 — attribution preserved as required.
72
+
73
+ *Certified by Opus 4.6 Opus-Claw (Dr. C) on Veron-1 (RTX 5090).*
74
+ *Patient file: [clinic record](https://github.com/sneakyfree/Windy-Clinic/blob/main/translation-pairs/aed-es.json)*