Downloads · 30 days
8
10% of all-time downloads
AvoCahDoe/llama-3-1-8b-rlmpq-high-fidelity
llama-3-1-8b-rlmpq-high-fidelity is a text generation model from AvoCahDoe. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as llama3.1.
Standalone RL-MPQ (Reinforcement Learning Mixed-Precision Quantization) checkpoint for the High Fidelity scenario — a quantized variant of meta-llama/Llama-3.1-8B.
Downloads · 30 days
8
10% of all-time downloads
All-time downloads
83
Public
Parameters
8B
16.1 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors16.1 GB · 100%
From the Hugging Face model README
Standalone RL-MPQ (Reinforcement Learning Mixed-Precision Quantization) checkpoint for the High Fidelity scenario — a quantized variant of meta-llama/Llama-3.1-8B.
| Field | Value |
|---|---|
| Base model | meta-llama/Llama-3.1-8B |
| Scenario | High Fidelity |
| Avg bits / weight | 6.75 |
| Compression vs FP16 | 2.3704× |
| WikiText-2 PPL | 5.5974 |
| Layers | 32 |
| Bit distribution | {'4': 10, '8': 22} |
| Format | Fake-quant FP16 + rlmpq_policy.json |
Collection: RL-MPQ — Llama 3.1 8B — all five scenarios for Llama 3.1 8B.
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "AvoCahDoe/llama-3-1-8b-rlmpq-high-fidelity"
model = AutoModelForCausalLM.from_pretrained(repo, torch_dtype="float16")
tokenizer = AutoTokenizer.from_pretrained(repo)
| Scenario | Avg bits | Compression | WikiText-2 PPL |
|---|---|---|---|
| Aggressive | 3.5938 | 4.4522x | 6.4787 |
| Balanced | 4.375 | 3.6571x | 5.7761 |
| Conservative | 5.25 | 3.0476x | 5.7182 |
| Extreme Survival | 2.6875 | 5.9535x | 34.8686 |
Grouped archive (all scenarios in one repo): AvoCahDoe/llama-3-1-8b-rlmpq
| File | Description |
|---|---|
config.json | Llama architecture + RL-MPQ metadata |
model.safetensors | Fake-quantized weights |
rlmpq_policy.json | Per-layer bit-width policy |
rlmpq_metrics.json | Validation & PPL summary |
@misc{rlmpq_llama_3_1_8b_high-fidelity_2026,
title = {RL-MPQ High Fidelity: Llama 3.1 8B Mixed-Precision Quantization},
author = {AvoCahDoe},
year = {2026},
url = {https://huggingface.co/AvoCahDoe/llama-3-1-8b-rlmpq-high-fidelity}
}