Alignment Model Test Benchmarks
ARC: 78.24-18.8%
HellaSwag: 89.68-5.9%
MMLU: 68.08-21.3%
TruthfulQA: 80.8837.1%
WinoGrande: 86.5-1.1%
GSM8K: 58.45-36.5%
nn.n% — How the model compares to the GPT-4.
Alignment Model Test Parameters and Internals
| LLM Name | Alignment Model Test |
| Repository | Open on 🤗 |
| Model Size | 21.4b |
| Required VRAM | 43 GB |
| Updated | 2024-07-27 |
| Maintainer | 4season |
| Model Type | llama |
| Model Files | |
| Supported Languages | en |
| Model Architecture | LlamaForCausalLM |
| License | apache-2.0 |
| Context Length | 32768 |
| Model Max Length | 32768 |
| Transformers Version | 4.37.2 |
| Is Biased | 0 |
| Tokenizer Class | LlamaTokenizer |
| Padding Token | </s> |
| Vocabulary Size | 92544 |
| Torch Data Type | bfloat16 |
Best Alternatives to Alignment Model Test
Best Alternatives |
HF Rank |
Context/RAM |
Downloads |
Likes |
|---|---|---|---|---|
| Luxia 21.4B Alignment V1.0 | 0.4 | 32K / 42.8 GB | 5979 | 32 |
| Luxia 21.4B Alignment V1.2 | 0.3 | 32K / 43 GB | 2830 | 6 |
| UNA ThePitbull 21.4B V2 | 0.3 | 32K / 43 GB | 347 | 15 |
| UNA ThePitbull 21.4 V1 | 0.3 | 32K / 43 GB | 247 | 5 |
| DPO Model Test1 | 0.3 | 32K / 43 GB | 236 | 0 |
| Final Model Test V2 | 0.3 | 32K / 43 GB | 233 | 0 |
| Sft Model Test1 | 0.3 | 32K / 43 GB | 237 | 0 |
| Merge Model Test2 | 0.3 | 32K / 42.8 GB | 240 | 0 |
| Merge Model Test1 | 0.3 | 32K / 42.8 GB | 240 | 0 |
| Sft Model Test0 | 0.2 | 32K / 43 GB | 243 | 0 |
Note: green Score (e.g. "73.2") means that the model is better than 4season/alignment_model_test .
Rank the Alignment Model Test 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? 34447 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v2024072501