chat-EOS fix: eos_token -> <turn|> (clean stop in generic apps) 5623d64 verified mlboydaisuke commited on 11 days ago
chat-EOS fix: eos_token -> <turn|> (clean stop in generic apps) 2c38bf9 verified mlboydaisuke commited on 11 days ago
Gemma 4 12B: drop simple-kernel gemma4_12b_qat_decode_int8lin_msdpa (superseded by _g8) d619e32 verified mlboydaisuke commited on 12 days ago
Gemma 4 12B: gemma4_12b_qat_decode_int4linsym_msdpa_g8 (higher-occupancy decode bundle) 34ccf9a verified mlboydaisuke commited on 12 days ago
Gemma 4 12B: gemma4_12b_qat_decode_int8lin_msdpa_g8 (higher-occupancy decode bundle) d1f03bb verified mlboydaisuke commited on 12 days ago
Gemma 4 12B: card -> higher-occupancy (_g8) kernel as the ship 24ded10 verified mlboydaisuke commited on 12 days ago
Gemma 4 12B: gemma4_12b_qat_decode_int8lin_msdpa (metal-sdpa decode bundle) 99f1dbb verified mlboydaisuke commited on 12 days ago