-
Add tokenizer_config.json from IT repo
-
Add tokenizer.json from IT repo
-
Update weights: 10B token continue-pretraining on finepdfs eng_Latn (val ppl 36.8 → 32.3)
-
Fix weights: strip _orig_mod. prefix (torch.compile artifact) so model loads without workarounds
-
Fix generate.py: correct repo name and add HF usage example
-
Add generation_config.json for default generation settings
-
Update config.json with auto_map and HF fields
-
Rewrite modeling_eve.py with HF-compatible EveMoEForCausalLM
-
Add HuggingFace transformers integration (AutoModelForCausalLM support)
-
Upload modeling_eve.py with huggingface_hub
-
Upload modeling_eve.py with huggingface_hub
-
Upload model.safetensors with huggingface_hub
-
Upload config.json with huggingface_hub
-
Upload modeling_eve.py with huggingface_hub
-
Upload configuration_eve.py with huggingface_hub
-
Update README.md
-
Model card.
-
Upload folder using huggingface_hub
-
initial commit
