vllm (pretrained=/root/autodl-tmp/Apriel-Nemotron-15b-Thinker,add_bos_token=true,max_model_len=8096,dtype=bfloat16), gen_kwargs: (None), limit: 250.0, num_fewshot: 5, batch_size: auto

Tasks Version Filter n-shot Metric Value Stderr
gsm8k 3 flexible-extract 5 exact_match ↑ 0.808 ± 0.0250
strict-match 5 exact_match ↑ 0.796 ± 0.0255

vllm (pretrained=/root/autodl-tmp/Apriel-Nemotron-15b-Thinker,add_bos_token=true,max_model_len=3096,dtype=bfloat16), gen_kwargs: (None), limit: 500.0, num_fewshot: 5, batch_size: auto

Tasks Version Filter n-shot Metric Value Stderr
gsm8k 3 flexible-extract 5 exact_match ↑ 0.798 ± 0.0180
strict-match 5 exact_match ↑ 0.804 ± 0.0178

vllm (pretrained=/root/autodl-tmp/Apriel-Nemotron-15b-Thinker,add_bos_token=true,max_model_len=3048,dtype=bfloat16), gen_kwargs: (None), limit: 15.0, num_fewshot: None, batch_size: auto

Groups Version Filter n-shot Metric Value Stderr
mmlu 2 none acc ↑ 0.6959 ± 0.0150
- humanities 2 none acc ↑ 0.7231 ± 0.0299
- other 2 none acc ↑ 0.7026 ± 0.0318
- social sciences 2 none acc ↑ 0.7722 ± 0.0300
- stem 2 none acc ↑ 0.6246 ± 0.0278

vllm (pretrained=/root/autodl-tmp/80-128,add_bos_token=true,max_model_len=8096,dtype=bfloat16), gen_kwargs: (None), limit: 250.0, num_fewshot: 5, batch_size: auto

Tasks Version Filter n-shot Metric Value Stderr
gsm8k 3 flexible-extract 5 exact_match ↑ 0.812 ± 0.0248
strict-match 5 exact_match ↑ 0.796 ± 0.0255

vllm (pretrained=/root/autodl-tmp/80-128,add_bos_token=true,max_model_len=3096,dtype=bfloat16), gen_kwargs: (None), limit: 500.0, num_fewshot: 5, batch_size: auto

Tasks Version Filter n-shot Metric Value Stderr
gsm8k 3 flexible-extract 5 exact_match ↑ 0.802 ± 0.0178
strict-match 5 exact_match ↑ 0.798 ± 0.0180

vllm (pretrained=/root/autodl-tmp/80-128,add_bos_token=true,max_model_len=3048,dtype=bfloat16), gen_kwargs: (None), limit: 15.0, num_fewshot: None, batch_size: auto

Groups Version Filter n-shot Metric Value Stderr
mmlu 2 none acc ↑ 0.6959 ± 0.0151
- humanities 2 none acc ↑ 0.7128 ± 0.0303
- other 2 none acc ↑ 0.7077 ± 0.0318
- social sciences 2 none acc ↑ 0.7500 ± 0.0312
- stem 2 none acc ↑ 0.6421 ± 0.0273

vllm (pretrained=/root/autodl-tmp/80-256,add_bos_token=true,max_model_len=8096,dtype=bfloat16), gen_kwargs: (None), limit: 250.0, num_fewshot: 5, batch_size: auto

Tasks Version Filter n-shot Metric Value Stderr
gsm8k 3 flexible-extract 5 exact_match ↑ 0.848 ± 0.0228
strict-match 5 exact_match ↑ 0.820 ± 0.0243

vllm (pretrained=/root/autodl-tmp/80-256,add_bos_token=true,max_model_len=3096,dtype=bfloat16), gen_kwargs: (None), limit: 500.0, num_fewshot: 5, batch_size: auto

Tasks Version Filter n-shot Metric Value Stderr
gsm8k 3 flexible-extract 5 exact_match ↑ 0.838 ± 0.0165
strict-match 5 exact_match ↑ 0.828 ± 0.0169

vllm (pretrained=/root/autodl-tmp/80-256,add_bos_token=true,max_model_len=3048,dtype=bfloat16), gen_kwargs: (None), limit: 15.0, num_fewshot: None, batch_size: auto

Groups Version Filter n-shot Metric Value Stderr
mmlu 2 none acc ↑ 0.7018 ± 0.0149
- humanities 2 none acc ↑ 0.7333 ± 0.0296
- other 2 none acc ↑ 0.6923 ± 0.0322
- social sciences 2 none acc ↑ 0.7556 ± 0.0304
- stem 2 none acc ↑ 0.6526 ± 0.0271
Downloads last month
1
Safetensors
Model size
15B params
Tensor type
BF16
·
I8
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for noneUsername/Apriel-Nemotron-15b-Thinker-W8A8

Quantized
(11)
this model