llama-3.1#

  • Context Length: 131072

  • Model Name: llama-3.1

  • Languages: en, de, fr, it, pt, hi, es, th

  • Abilities: generate

  • Description: Llama 3.1 is an auto-regressive language model that uses an optimized transformer architecture

Specifications#

Model Spec 1 (pytorch, 8 Billion)#

  • Model Format: pytorch

  • Model Size (in billions): 8

  • Quantizations: 4-bit, 8-bit, none

  • Engines: vLLM, Transformers, SGLang (vLLM and SGLang only available for quantization none)

  • Model ID: meta-llama/Meta-Llama-3.1-8B

  • Model Hubs: Hugging Face, ModelScope

Execute the following command to launch the model, remember to replace ${quantization} with your chosen quantization method from the options listed above:

xinference launch --model-engine ${engine} --model-name llama-3.1 --size-in-billions 8 --model-format pytorch --quantization ${quantization}

Model Spec 2 (ggufv2, 8 Billion)#

  • Model Format: ggufv2

  • Model Size (in billions): 8

  • Quantizations: Q2_K, Q3_K_L, Q3_K_M, Q3_K_S, Q4_0, Q4_1, Q4_K_M, Q4_K_S, Q5_0, Q5_1, Q5_K_M, Q5_K_S, Q6_K, Q8_0

  • Engines: llama.cpp

  • Model ID: QuantFactory/Meta-Llama-3.1-8B-GGUF

  • Model Hubs: Hugging Face

Execute the following command to launch the model, remember to replace ${quantization} with your chosen quantization method from the options listed above:

xinference launch --model-engine ${engine} --model-name llama-3.1 --size-in-billions 8 --model-format ggufv2 --quantization ${quantization}

Model Spec 3 (pytorch, 70 Billion)#

  • Model Format: pytorch

  • Model Size (in billions): 70

  • Quantizations: 4-bit, 8-bit, none

  • Engines: vLLM, Transformers, SGLang (vLLM and SGLang only available for quantization none)

  • Model ID: meta-llama/Meta-Llama-3.1-70B

  • Model Hubs: Hugging Face, ModelScope

Execute the following command to launch the model, remember to replace ${quantization} with your chosen quantization method from the options listed above:

xinference launch --model-engine ${engine} --model-name llama-3.1 --size-in-billions 70 --model-format pytorch --quantization ${quantization}

Model Spec 4 (pytorch, 405 Billion)#

  • Model Format: pytorch

  • Model Size (in billions): 405

  • Quantizations: 4-bit, 8-bit, none

  • Engines: vLLM, Transformers, SGLang (vLLM and SGLang only available for quantization none)

  • Model ID: meta-llama/Meta-Llama-3.1-405B

  • Model Hubs: Hugging Face, ModelScope

Execute the following command to launch the model, remember to replace ${quantization} with your chosen quantization method from the options listed above:

xinference launch --model-engine ${engine} --model-name llama-3.1 --size-in-billions 405 --model-format pytorch --quantization ${quantization}