Autotrain compatible Base model:finetune:qwen/qwen2... Base model:qwen/qwen2.5-14b Conversational Dataset:allura-org/celeste-1.x... Dataset:anthracite-org/kalo-op... Dataset:cognitivecomputations/... Dataset:epiculous/synthrp-gens... Dataset:epiculous/synthstruct-... Dataset:gryphe/chatgpt-4o-writ... Dataset:gryphe/sonnet3.5-charc... Dataset:gryphe/sonnet3.5-slimo... Dataset:nopm/opus writingstruc... Dataset:nothingiisreal/reddit-... Endpoints compatible Generated from trainer Instruct Qwen2 Region:us Safetensors Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2
EVA Qwen2.5 14B V0.2 Benchmarks
MMLU Pro: 45.95
GPQA: 19.24
MUSR: 19.63
BBH: 43.61
IFEval: 40.38 vs 88 (so35)-54.1%
MATH Lvl 5: 33.84
LLME Score: 0.36253
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)
EVA Qwen2.5 14B V0.2 Parameters and Internals
| LLM Name | EVA Qwen2.5 14B V0.2 |
| Repository 🤗 | https://huggingface.co/EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2 |
| Base Model(s) | |
| Model Size | 14b |
| Required VRAM | 29.7 GB |
| Updated | 2025-01-19 |
| Maintainer | EVA-UNIT-01 |
| Model Type | qwen2 |
| Instruction-Based | Yes |
| Model Files | |
| Model Architecture | Qwen2ForCausalLM |
| License | apache-2.0 |
| Context Length | 131072 |
| Model Max Length | 131072 |
| Transformers Version | 4.45.1 |
| Tokenizer Class | Qwen2Tokenizer |
| Padding Token | <|endoftext|> |
| Vocabulary Size | 152064 |
| Torch Data Type | bfloat16 |
| Errors | replace |
Best Alternatives to EVA Qwen2.5 14B V0.2
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Lamarck 14B V0.6 | 128K / 29.7 GB | 643 | 11 |
| Qwen2.5 14B Vimarckoso V3 | 128K / 29.7 GB | 769 | 10 |
| Qwenvergence 14B V9 | 128K / 29.7 GB | 12 | 3 |
| Q2.5 Veltha 14B 0.5 | 128K / 29.7 GB | 354 | 9 |
| Rombos LLM V2.6 Qwen 14B | 128K / 29.7 GB | 7464 | 49 |
| Q2.5 Veltha 14B | 128K / 29.7 GB | 137 | 6 |
| Qwenvergence 14B V3 Prose | 128K / 29.7 GB | 228 | 2 |
| Ultiima 14B | 128K / 29.7 GB | 11 | 1 |
| Fusion4 14B Instruct | 128K / 29.7 GB | 52 | 0 |
| Miscii 14B 1028 | 128K / 29.6 GB | 136 | 17 |
Note: green Score (e.g. "73.2") means that the model is better than EVA-UNIT-01/EVA-Qwen2.5-14B-v0.2 .
Rank the EVA Qwen2.5 14B V0.2 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? 41636 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227