Fathom: upload expert-e1-static/tokenizer_config.json

Fathom: upload expert-e1-static/tokenizer.model

Fathom: upload expert-e1-static/tokenizer.json

Fathom: upload expert-e1-static/special_tokens_map.json

Fathom: upload expert-e1-static/chat_template.jinja

Fathom: upload expert-e1-static/adapter_model.safetensors

Fathom: upload expert-e1-static/adapter_config.json

Fathom: upload unified-v2/training_log.json

Fathom: upload unified-v2/tokenizer_config.json

Fathom: upload unified-v2/tokenizer.json

Fathom: upload unified-v2/chat_template.jinja

Fathom: upload unified-v2/adapter_model.safetensors

Fathom: upload unified-v2/adapter_config.json

Fathom: update model card with full benchmark results

Update model card with full Plan A details and results

Plan A LoRA adapter: Mixtral-8x7B cybersecurity fine-tune (3000 steps, QLoRA r=32)

initial commit