Upload tokenizer_config.json with huggingface_hub

Delete chat_template.jinja with huggingface_hub

Delete processor_config.json with huggingface_hub

Delete preprocessor_config.json with huggingface_hub

Delete model-00016-of-00016.safetensors with huggingface_hub

Delete model-00015-of-00016.safetensors with huggingface_hub

Delete model-00014-of-00016.safetensors with huggingface_hub

Delete model-00013-of-00016.safetensors with huggingface_hub

Delete model-00012-of-00016.safetensors with huggingface_hub

Delete model-00011-of-00016.safetensors with huggingface_hub

Delete model-00010-of-00016.safetensors with huggingface_hub

Delete model-00009-of-00016.safetensors with huggingface_hub

Delete model-00008-of-00016.safetensors with huggingface_hub

Delete model-00007-of-00016.safetensors with huggingface_hub

Delete model-00006-of-00016.safetensors with huggingface_hub

Delete model-00005-of-00016.safetensors with huggingface_hub

Delete model-00004-of-00016.safetensors with huggingface_hub

Delete model-00003-of-00016.safetensors with huggingface_hub

Delete model-00002-of-00016.safetensors with huggingface_hub

Delete model-00001-of-00016.safetensors with huggingface_hub

Upload merged Qwen3.5-35B-A3B + sparse moderation LoRA B400 (tinker-cookbook merge)

Reshard to 5GB shards for faster parallel download

Remove old shard model-00002-of-00002.safetensors

Remove old shard model-00001-of-00002.safetensors

Fix tokenizer (TokenizersBackend→correct class) + add model card

Upload merged Qwen3.5-35B-A3B + sparse moderation LoRA B400 (bf16)

initial commit