Llama-3.1-8B-Instruct-FP8
Support summary
Llama-3.1-8B-Instruct-FP8 is a checkpoint published by NVIDIA, built on the LlamaForCausalLM architecture. The newest indexed releases SGLang 0.5.19 and TensorRT-LLM 1.2.1 both document support. Official recipes or cookbooks cover vLLM, without pinning a stable release.
SGLang documents support in 2 of 13 indexed releases, from 0.5.18 to 0.5.19. TensorRT-LLM documents support in 2 of 2 indexed releases, from 1.2.0 to 1.2.1.
- Documented quantization
- FP8
- Configured context
- 131,072 tokens
- License
- llama3.1
Hardware, quantization and features count only statements about this checkpoint; engine-wide capabilities are listed with each release.
Hugging Face model card
- Publisher
- NVIDIA
- Task
- text generation
- License
- llama3.1
- Library
- transformers
- Architecture
- LlamaForCausalLM
- Weight tensor elements
- 8B
- Tensor types
- BF16, F8_E4M3
- Configured context
- 131,072 tokens
- Base models
- meta-llama/Llama-3.1-8B-Instruct
- Weight access
- Public
- Repository created
- 2024-08-29
Model notes and sources
Exact ModelOpt FP8 examples exist in SGLang and TensorRT-LLM; inspect hardware and build requirements.
- SGLang v0.5.19 · docs/docs/advanced_features/quantization.mdx · 2 indexed revisions, v0.5.18 to v0.5.19
- TensorRT-LLM v1.2.1 · docs/source/features/quantization.md · 2 indexed revisions, v1.2.0 to v1.2.1
- nvidia/Llama-3.1-8B-Instruct-FP8 publisher metadata
- nvidia/Llama-3.1-8B-Instruct-FP8 · publisher model card
- nvidia/Llama-3.1-8B-Instruct-FP8 · Hugging Face metadata
- nvidia/Llama-3.1-8B-Instruct-FP8 · publisher configuration
- vllm · models/meta-llama/Llama-3.1-8B-Instruct.yaml
- Reviewed
Engine support
3 engines
| Engine | Release | Checkpoint support | Evidence scope | Last known supported |
|---|---|---|---|---|
| vLLM | Version not pinnedRecipe / cookbook | Documented recipeSource: vllm · models/meta-llama/Llama-3.1-8B-Instruct.yamlReviewed | No supported release indexed | |
Features, dependencies and evidence for vLLM 0.28.0 | ||||
| SGLang | 0.5.19Newest indexed | Checkpoint evidenceSource: SGLang v0.5.19 · docs/docs/advanced_features/quantization.mdxReviewed | 0.5.19This release | |
Features, dependencies and evidence for SGLang 0.5.19 | ||||
| TensorRT-LLM | 1.2.1Newest indexed | Checkpoint evidenceSource: TensorRT-LLM v1.2.1 · docs/source/features/quantization.mdReviewed | 1.2.1This release | |
Features, dependencies and evidence for TensorRT-LLM 1.2.1 | ||||