GPT4 X Alpasta 30B 4bit is an open-source language model by MetaIX. Features: 30b LLM, VRAM: 16.9GB, Context: 2K, Quantized, HF Score: 60.8, LLM Explorer Score: 0.1, Arc: 63.1, HellaSwag: 83.6, MMLU: 57.7, TruthfulQA: 51.5, WinoGrande: 78.2, GSM8K: 30.5.
Model Card on HF 🤗: https://huggingface.co/MetaIX/GPT4-X-Alpasta-30b-4bit
GPT4 X Alpasta 30B 4bit Benchmarks
ARC: 63.05 vs 96.7 (so35)-34.8%
HellaSwag: 83.56 vs 95.3 (gpt4)-12.3%
MMLU: 57.71 vs 88.3 (so35)-34.6%
TruthfulQA: 51.52 vs 59 (gpt4)-12.7%
WinoGrande: 78.22 vs 87.5 (gpt4)-10.6%
GSM8K: 30.48 vs 96.4 (so35)-68.4%
LLME Score: 0.10017
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)
GPT4 X Alpasta 30B 4bit Parameters and Internals
| Model Type |
| ||||||||||
| Additional Notes |
| ||||||||||
| Release Notes |
|
| LLM Name | GPT4 X Alpasta 30B 4bit |
| Repository 🤗 | https://huggingface.co/MetaIX/GPT4-X-Alpasta-30b-4bit |
| Base Model(s) | |
| Model Size | 30b |
| Required VRAM | 16.9 GB |
| Updated | 2025-11-03 |
| Maintainer | MetaIX |
| Model Type | llama |
| Model Files | |
| GGML Quantization | Yes |
| Quantization Type | ggml|q4|4bit |
| Model Architecture | LlamaForCausalLM |
| Context Length | 2048 |
| Model Max Length | 2048 |
| Transformers Version | 4.29.0.dev0 |
| Tokenizer Class | LlamaTokenizer |
| Vocabulary Size | 32016 |
| Torch Data Type | float16 |
Best Alternatives to GPT4 X Alpasta 30B 4bit
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Alpacino30b | 2K / 64.1 GB | 522 | 68 |
| GPT4 X Alpaca 30B 4bit | 2K / 16.9 GB | 1927 | 162 |
| Tamillama Tiny 30m | 2K / 0.1 GB | 31 | 15 |
| Bluemoonrp 30B | 2K / 17.5 GB | 781 | 24 |
| OpenAssistant Llama 30B 4bit | 2K / 16.9 GB | 5 | 69 |
| Tenebra 30B Alpha01 FP16 | 16K / 65 GB | 1833 | 9 |
| Tenebra 30B Alpha01 4BIT | 16K / 19.4 GB | 65 | 1 |
| ...nebra 30B Alpha01 EXL2 2 80bpw | 16K / 11.9 GB | 3 | 1 |
| Tenebra 30B Alpha01 EXL2 3bpw | 16K / 12.7 GB | 6 | 0 |
| ...nebra 30B Alpha01 EXL2 2 50bpw | 16K / 10.7 GB | 5 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than MetaIX/GPT4-X-Alpasta-30b-4bit .
Rank the GPT4 X Alpasta 30B 4bit 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? 52473 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a