Yz

qwen2

Qwen2 is a new series of large language models from Alibaba group

Araç Kullanımı 0.5b 1.5b 7b 72b
Hızlı Kurulum (Ollama kuruluysa) 
ollama run qwen2

Ollama kurulu değil mi? ollama.com/download — Windows, macOS ve Linux için ücretsiz. İlk çalıştırmada model indirilir, sonrası tamamen çevrimdışıdır.

Varyantlar

Boyut büyüdükçe kalite artar, donanım ihtiyacı yükselir. Başlangıç için küçük varyantı deneyin.

EtiketBoyutBağlamGirdiKomut
latest 4.4GB 32K Text ollama run qwen2:latest
0.5b 352MB 32K Text ollama run qwen2:0.5b
1.5b 935MB 32K Text ollama run qwen2:1.5b
7b 4.4GB 32K Text ollama run qwen2:7b
72b 41GB 32K Text ollama run qwen2:72b
0.5b-instruct 352MB 32K Text ollama run qwen2:0.5b-instruct
0.5b-instruct-q2_K 339MB 32K Text ollama run qwen2:0.5b-instruct-q2_K
0.5b-instruct-q3_K_S 338MB 32K Text ollama run qwen2:0.5b-instruct-q3_K_S
0.5b-instruct-q3_K_M 355MB 32K Text ollama run qwen2:0.5b-instruct-q3_K_M
0.5b-instruct-q3_K_L 369MB 32K Text ollama run qwen2:0.5b-instruct-q3_K_L
0.5b-instruct-q4_0 352MB 32K Text ollama run qwen2:0.5b-instruct-q4_0
0.5b-instruct-q4_1 375MB 32K Text ollama run qwen2:0.5b-instruct-q4_1
0.5b-instruct-q4_K_S 385MB 32K Text ollama run qwen2:0.5b-instruct-q4_K_S
0.5b-instruct-q4_K_M 398MB 32K Text ollama run qwen2:0.5b-instruct-q4_K_M
0.5b-instruct-q5_0 397MB 32K Text ollama run qwen2:0.5b-instruct-q5_0
0.5b-instruct-q5_1 419MB 32K Text ollama run qwen2:0.5b-instruct-q5_1
0.5b-instruct-q5_K_S 413MB 32K Text ollama run qwen2:0.5b-instruct-q5_K_S
0.5b-instruct-q5_K_M 420MB 32K Text ollama run qwen2:0.5b-instruct-q5_K_M
0.5b-instruct-q6_K 506MB 32K Text ollama run qwen2:0.5b-instruct-q6_K
0.5b-instruct-q8_0 531MB 32K Text ollama run qwen2:0.5b-instruct-q8_0
0.5b-instruct-fp16 994MB 32K Text ollama run qwen2:0.5b-instruct-fp16
1.5b-instruct 935MB 32K Text ollama run qwen2:1.5b-instruct
1.5b-instruct-q2_K 676MB 32K Text ollama run qwen2:1.5b-instruct-q2_K
1.5b-instruct-q3_K_S 761MB 32K Text ollama run qwen2:1.5b-instruct-q3_K_S
1.5b-instruct-q3_K_M 824MB 32K Text ollama run qwen2:1.5b-instruct-q3_K_M

72 varyant daha var — tümü için ollama.com.

Model Detayları ve Benchmarklar (kaynak: ollama.com)

Qwen2 is trained on data in 29 languages, including English and Chinese.

It is available in 4 parameter sizes: 0.5B, 1.5B, 7B, 72B.

In the 7B and 72B models, context length has been extended to 128k tokens.

Models Qwen2-0.5B Qwen2-1.5B Qwen2-7B Qwen2-72B
Params 0.49B 1.54B 7.07B 72.71B
Non-Emb Params 0.35B 1.31B 5.98B 70.21B
GQA True True True True
Tie Embedding True True False False
Context Length 32K 32K 128K 128K

Supported languages

This is in addition to English and Chinese

Regions Languages
Western Europe German, French, Spanish, Portuguese, Italian, Dutch
Eastern & Central Europe Russian, Czech, Polish
Middle East Arabic, Persian, Hebrew, Turkish
Eastern Asia Japanese, Korean
South-Eastern Asia Vietnamese, Thai, Indonesian, Malay, Lao, Burmese, Cebuano, Khmer, Tagalog
Southern Asia Hindi, Bengali, Urdu

Performance

image.png

image.png

image.png

image.png

License

All models with the exception of Qwen2 72B (both instruct and base models) are Apache 2.0 licensed.

Qwen2 72B model still uses the original Qianwen License.