YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Merged from all available midtraining checkpoints using mergekit-evolve, optimizing for arc_easy and arc_challenge.

Tasks Version Filter n-shot Metric Value Stderr
arc_challenge 1 none 0 acc ↑ 0.2056 ± 0.0118
none 0 acc_norm ↑ 0.2415 ± 0.0125
arc_easy 1 none 0 acc ↑ 0.5063 ± 0.0103
none 0 acc_norm ↑ 0.4848 ± 0.0103
hellaswag 1 none 0 acc ↑ 0.2901 ± 0.0045
none 0 acc_norm ↑ 0.3173 ± 0.0046
piqa 1 none 0 acc ↑ 0.6224 ± 0.0113
none 0 acc_norm ↑ 0.6148 ± 0.0114
BananaMind Base Bench 1.1
Overall Elo: 994
Accuracy: 179/350 (51.14%)
Weighted accuracy: 47.20%
language_completion: Elo 1278 | 46/50 (92.00%) | weighted 90.45%
commonsense: Elo 1022 | 32/50 (64.00%) | weighted 59.31%
world_knowledge: Elo 1040 | 32/50 (64.00%) | weighted 61.76%
context_tracking: Elo 824 | 14/50 (28.00%) | weighted 27.39%
quantitative: Elo 828 | 11/50 (22.00%) | weighted 22.40%
logical_reasoning: Elo 985 | 20/50 (40.00%) | weighted 34.80%
code_completion: Elo 1076 | 24/50 (48.00%) | weighted 47.45%
====================================================================
  /root/best_midtrain_merge (149,935,616 params) RESULTS
====================================================================
  Raw continuation accuracy        40.00%
  Length-normalized accuracy       40.00%
  Primary (acc_norm)           40.00%
====================================================================
Downloads last month
26
Safetensors
Model size
0.1B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support