Arxiv:2305.18290 Autotrain compatible Base model:finetune:qwen/qwen2... Base model:qwen/qwen2-0.5b-ins... Conversational Dataset:trl-lib/capybara-prefe... Dpo Endpoints compatible Generated from trainer Instruct Qwen2 Region:us Safetensors Trl
DPO Qwen2 Benchmarks
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)
DPO Qwen2 Parameters and Internals
| LLM Name | DPO Qwen2 |
| Repository 🤗 | https://huggingface.co/trl-lib/dpo-qwen2 |
| Model Name | dpo-qwen2 |
| Base Model(s) | |
| Model Size | 0.5b |
| Required VRAM | 2 GB |
| Updated | 2024-10-01 |
| Maintainer | trl-lib |
| Model Type | qwen2 |
| Instruction-Based | Yes |
| Model Files | |
| Model Architecture | Qwen2ForCausalLM |
| Context Length | 32768 |
| Model Max Length | 32768 |
| Transformers Version | 4.45.0.dev0 |
| Tokenizer Class | Qwen2Tokenizer |
| Padding Token | <|endoftext|> |
| Vocabulary Size | 151936 |
| Torch Data Type | float32 |
| Errors | replace |
Best Alternatives to DPO Qwen2
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Qwen2 0.5B Abyme Merge3 | 128K / 1.3 GB | 7 | 0 |
| Qwen2 0.5B Abyme Merge2 | 128K / 0.3 GB | 5 | 0 |
| Qwen2.5 0.5B Instruct | 32K / 1 GB | 48331 | 50 |
| Qwen2 0.5B Instruct | 32K / 1 GB | 278170 | 155 |
| Replete LLM V2.5 Qwen 0.5B | 32K / 1.3 GB | 180 | 2 |
| Qwen2.5 0.5B Test Ft | 32K / 1 GB | 99 | 3 |
| Qwen2.5 0.5B Instruct | 32K / 1 GB | 417 | 1 |
| Qwen2 0.5B DPO | 32K / 2 GB | 56 | 2 |
| Qwen2.5 0.5B Instruct | 32K / GB | 32 | 2 |
| Qwama 0.5B Instruct | 32K / 0.9 GB | 1192 | 13 |
Note: green Score (e.g. "73.2") means that the model is better than trl-lib/dpo-qwen2 .
Rank the DPO Qwen2 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? 36542 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v2024072803