Ba2han_-_Pretraining_Turkish-test-5500s-exl2

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Pretraining_Turkish-test-5500s - EXL2

• Model creator: https://huggingface.co/Ba2han/
• Original model: https://huggingface.co/Ba2han/Pretraining_Turkish-test-5500s/

Available sizes

| Branch | Bits | Description | | ----- | ---- | ------- | ------ | ------ | ------ | ------ | ------------ | | 8_0 | 8.0 | Maximum quality that ExLlamaV2 can produce, near unquantized performance. | | 6_5 | 6.5 | Very similar to 8.0, good tradeoff of size vs performance, recommended. | | 5_0 | 5.0 | Slightly lower quality vs 6.5, but usable on 8GB cards. | | 4_25 | 4.25 | GPTQ equivalent bits per weight, slightly higher quality. | | 3_5 | 3.5 | Lower quality, only use if you have to. |

Download instructions

With git:

git clone --single-branch --branch 6_5 https://huggingface.co/Ba2han_-_Pretraining_Turkish-test-5500s-exl2 Pretraining_Turkish-test-5500s-6_5

With huggingface hub:

pip3 install huggingface-hub

To download a specific branch, use the --revision parameter. For example, to download the 6.5 bpw branch: Linux:

huggingface-cli download Ba2han_-_Pretraining_Turkish-test-5500s-exl2 --revision 6_5 --local-dir Pretraining_Turkish-test-5500s-6_5 --local-dir-use-symlinks False

Windows (which apparently doesn't like _ in folders sometimes?):

huggingface-cli download Ba2han_-_Pretraining_Turkish-test-5500s-exl2 --revision 6_5 --local-dir Pretraining_Turkish-test-5500s-6.5 --local-dir-use-symlinks False

Original model description:

license: mit language:

• tr library_name: transformers

---

Pretrained on 1B (mostly Turkish) tokens from HF and "high quality" scraped data using 1 RTX 3090. The training will continue. The model already can produce sensible sentences in Turkish.

---

HF kaynaklı ve scrape edilen 1 Milyar (çoğunlukla Türkçe) token ile 1 RTX 3090 kullanılarak eğitilmiştir. Eğitim devam edecek. Model şimdiden düzgün Türkçe cümleler kurabiliyor:

(top_k=24, repetition_penalty=1.1, temperature=0.12, seed=1022)