Chinasky71/translategemma-12b-it-subtitles
translation model by Chinasky71
Older snapshot: check the original source for current information.
From the publisher
Model card & documentation
Source previewRead the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.
Full-precision (bfloat16) merged weights of a TranslateGemma-12B model fine-tuned for English→Italian subtitle translation. It is the model used by Transcriber Pro to translate subtitles; the quantized GGUF versions (for 8–12 GB graphics cards and AMD/Vulkan) are in translategemma-12b-it-subtitles-gguf. In a blind judgement on 900 subtitle lines from two held-out films, the serious-error rate of this model run in 4-bit NF4 was 6.6%. The GGUF quantizations are non-inferior to it with a 3-point margin (Q4KM 6.2%, IQ3M 7.1%, Q3KM 7.4%). Caveats: one judge type, one kind of material (films), EN→IT only. See the project repository for the method. The model keeps the TranslateGemma chat format: a single user turn whose content carries the language codes and the text. The snippet follows the base model's usage; it was not run end to end on this checkpoint with…
Read the full model card ↗ · Preview checked 2026-11-01T02:54:38.992Z
Inside the original model card — Document outline
- TranslateGemma 12B fine-tuned for English→Italian subtitles
- Quality
- Usage
- Intended use and limits
- License and use
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Documentation belongs to its respective authors. Reported project/model license: gemma. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
Model overview
Chinasky71/translategemma-12b-it-subtitles is a translation model repository published by Chinasky71 on Hugging Face. The source reports the transformers library.
This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.
Model facts
- Task
- translation
- Library
- transformers
- Recent downloads (30 days)
- 0
- Cumulative likes
- 0
- Hugging Face trending score
- 0
- Reported safetensors parameters
- 12,187,325,040
- Architecture
- Gemma3ForConditionalGeneration
- License
- gemma
- Access
- Not gated by Hugging Face
- Created
- 2026-10-27T06:37:01.000Z
- Last modified
- 2026-10-27T06:45:22.000Z
Compatibility and lineage
Reported languages: en, it
Reported base models: google/translategemma-12b-it
Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.
Use and evaluate
Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.
No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.
Source and freshness
Source: Hugging Face Hub. Metadata observed 2026-10-27T06:47:19.891Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.