8bit Conversational En Facebook Instruct Llama Llama-3 Meta Mlx Pytorch Quantized Region:us Safetensors Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/mlx-community/Meta-Llama-3-8B-Instruct-8bit
Meta Llama 3 8B Instruct 8bit Benchmarks
ARC: 59.64 vs 96.7 (so35)-38.3%
HellaSwag: 78.84 vs 95.3 (gpt4)-17.3%
MMLU: 65.06 vs 88.3 (so35)-26.3%
TruthfulQA: 50.17 vs 59 (gpt4)-15%
WinoGrande: 74.98 vs 87.5 (gpt4)-14.3%
GSM8K: 57.32 vs 96.4 (so35)-40.5%
LLME Score: 0.20115
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Meta Llama 3 8B Instruct 8bit Parameters and Internals
| Model Type |
|
| LLM Name | Meta Llama 3 8B Instruct 8bit |
| Repository 🤗 | https://huggingface.co/mlx-community/Meta-Llama-3-8B-Instruct-8bit |
| Base Model(s) | |
| Model Size | 8b |
| Required VRAM | 9 GB |
| Updated | 2025-02-05 |
| Maintainer | mlx-community |
| Model Type | llama |
| Instruction-Based | Yes |
| Model Files | |
| Supported Languages | en |
| Quantization Type | 8bit |
| Model Architecture | LlamaForCausalLM |
| License | other |
| Context Length | 8192 |
| Model Max Length | 8192 |
| Transformers Version | 4.40.0 |
| Vocabulary Size | 128256 |
| Torch Data Type | bfloat16 |
Best Alternatives to Meta Llama 3 8B Instruct 8bit
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| ...B Instruct Gradient 1048K 4bit | 1024K / 4.5 GB | 31 | 2 |
| ...B Instruct Gradient 1048K 8bit | 1024K / 8.6 GB | 14 | 1 |
| ...truct Gradient 1048K Bpw6 EXL2 | 1024K / 6.7 GB | 6 | 2 |
| ...truct Gradient 1048K Bpw5 EXL2 | 1024K / 5.8 GB | 5 | 0 |
| Llama 3 8B Instruct 1048K 4bit | 1024K / 4.5 GB | 8 | 25 |
| Llama 3 8B Instruct 1048K 8bit | 1024K / 8.6 GB | 21 | 17 |
| ... Gradient 1048K 8.0bpw H8 EXL2 | 1024K / 8.6 GB | 5 | 3 |
| ...ct Gradient 1048K Bpw2.25 EXL2 | 1024K / 3.4 GB | 5 | 1 |
| Llama 3 8B Instruct 262K 2bit | 256K / 2.5 GB | 6 | 1 |
| ...B Instruct 262k V2 EXL2 5.0bpw | 256K / 5.8 GB | 5 | 1 |
Note: green Score (e.g. "73.2") means that the model is better than mlx-community/Meta-Llama-3-8B-Instruct-8bit .
Rank the Meta Llama 3 8B Instruct 8bit Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 42577 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227