Yz

granite3.1-moe

The IBM Granite 1B and 3B models are long-context mixture of experts (MoE) Granite models from IBM designed for low latency usage.

Araç Kullanımı 1b 3b
Hızlı Kurulum (Ollama kuruluysa) 
ollama run granite3.1-moe

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 2.0GB 128K Text ollama run granite3.1-moe:latest
1b 1.4GB 128K Text ollama run granite3.1-moe:1b
3b 2.0GB 128K Text ollama run granite3.1-moe:3b
1b-instruct-q2_K 524MB 128K Text ollama run granite3.1-moe:1b-instruct-q2_K
1b-instruct-q3_K_S 610MB 128K Text ollama run granite3.1-moe:1b-instruct-q3_K_S
1b-instruct-q3_K_M 671MB 128K Text ollama run granite3.1-moe:1b-instruct-q3_K_M
1b-instruct-q3_K_L 724MB 128K Text ollama run granite3.1-moe:1b-instruct-q3_K_L
1b-instruct-q4_0 781MB 128K Text ollama run granite3.1-moe:1b-instruct-q4_0
1b-instruct-q4_1 861MB 128K Text ollama run granite3.1-moe:1b-instruct-q4_1
1b-instruct-q4_K_S 787MB 128K Text ollama run granite3.1-moe:1b-instruct-q4_K_S
1b-instruct-q4_K_M 834MB 128K Text ollama run granite3.1-moe:1b-instruct-q4_K_M
1b-instruct-q5_0 941MB 128K Text ollama run granite3.1-moe:1b-instruct-q5_0
1b-instruct-q5_1 1.0GB 128K Text ollama run granite3.1-moe:1b-instruct-q5_1
1b-instruct-q5_K_S 941MB 128K Text ollama run granite3.1-moe:1b-instruct-q5_K_S
1b-instruct-q5_K_M 969MB 128K Text ollama run granite3.1-moe:1b-instruct-q5_K_M
1b-instruct-q6_K 1.1GB 128K Text ollama run granite3.1-moe:1b-instruct-q6_K
1b-instruct-q8_0 1.4GB 128K Text ollama run granite3.1-moe:1b-instruct-q8_0
1b-instruct-fp16 2.7GB 128K Text ollama run granite3.1-moe:1b-instruct-fp16
3b-instruct-q2_K 1.3GB 128K Text ollama run granite3.1-moe:3b-instruct-q2_K
3b-instruct-q3_K_S 1.5GB 128K Text ollama run granite3.1-moe:3b-instruct-q3_K_S
3b-instruct-q3_K_M 1.6GB 128K Text ollama run granite3.1-moe:3b-instruct-q3_K_M
3b-instruct-q3_K_L 1.8GB 128K Text ollama run granite3.1-moe:3b-instruct-q3_K_L
3b-instruct-q4_0 1.9GB 128K Text ollama run granite3.1-moe:3b-instruct-q4_0
3b-instruct-q4_1 2.1GB 128K Text ollama run granite3.1-moe:3b-instruct-q4_1
3b-instruct-q4_K_S 1.9GB 128K Text ollama run granite3.1-moe:3b-instruct-q4_K_S

8 varyant daha var — tümü için ollama.com.

Model Detayları ve Benchmarklar (kaynak: ollama.com)

Granite mixture of experts models

The IBM Granite 1B and 3B models are long-context mixture of experts (MoE) Granite models from IBM designed for low latency usage.

The models are trained on over 10 trillion tokens of data, the Granite MoE models are ideal for deployment in on-device applications or situations requiring instantaneous inference.

Parameter Sizes

1B:

ollama run granite3.1-moe:1b

3B:

ollama run granite3.1-moe:3b

Supported Languages

English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, Chinese (Simplified)

Capabilities

  • Summarization
  • Text classification
  • Text extraction
  • Question-answering
  • Retrieval Augmented Generation (RAG)
  • Code related tasks
  • Function-calling tasks
  • Multilingual dialog use cases
  • Long-context tasks including long document/meeting summarization, long document QA, etc.

Granite dense models

The Granite dense models are available in 2B and 8B parameter sizes designed to support tool-based use cases and for retrieval augmented generation (RAG), streamlining code generation, translation and bug fixing.

See model page

Learn more