Dan Saattrup Smart comments

Results 240 comments of


                                            Dan Saattrup Smart

[MODEL EVALUATION REQUEST] deepseek-ai/DeepSeek-R1-Distill-Qwen-32B

Live on the leaderboards 🎉

[MODEL EVALUATION REQUEST] deepseek-ai/DeepSeek-R1-Distill-Llama-70B

@Mikeriess According to [this vLLM issue](https://github.com/vllm-project/vllm/issues/12226), it should already be supported 🙂

[MODEL EVALUATION REQUEST] deepseek-ai/DeepSeek-R1-Distill-Llama-70B

@Mikeriess Thanks! 🙏 Results are live on the leaderboards now 🎉

[MODEL EVALUATION REQUEST] deepseek-ai/DeepSeek-R1-Distill-Llama-70B

Reopening due to #752.

[MODEL EVALUATION REQUEST] deepseek-ai/DeepSeek-R1-Distill-Llama-70B

Live on the leaderboards now 🎉

[MODEL EVALUATION REQUEST] EvaByte/EvaByte-Phase1

Not supported in vLLM yet.

[MODEL EVALUATION REQUEST] google/gemma-3-27b-it-qat-q4_0-gguf

@mathiasesn Thanks for adding them! Can I ask you to add the pretrained variants as well? :)

[MODEL EVALUATION REQUEST] Qwen/Qwen3-8B

Thanks @mathiasesn! It's live on the leaderboard now 🎉

[BENCHMARK DATASET REQUEST] The Danish Entailment Dataset

This dataset could be part of a new NLI task (https://github.com/EuroEval/EuroEval/issues/798), which can be put within the sequence classification task group.

European values evaluation

European Values Survey: https://europeanvaluesstudy.eu/methodology-data-documentation/data-downloads/