Compare commits
merge into: mirrors/karpathy--llm.c:main
mirrors/karpathy--llm.c:master
mirrors/karpathy--llm.c:main
mirrors/karpathy--llm.c:llama3
mirrors/karpathy--llm.c:feature/managed2
mirrors/karpathy--llm.c:gordicaleksa-fix_dataloader2
mirrors/karpathy--llm.c:feature/finetune_llama31py
mirrors/karpathy--llm.c:feature/restore_from_master
mirrors/karpathy--llm.c:refactor/code-to-zerocuh
mirrors/karpathy--llm.c:feature/gpt3v1
mirrors/karpathy--llm.c:fix/reshuffle_zero_grad
mirrors/karpathy--llm.c:overfit-batch-actually
mirrors/karpathy--llm.c:feature/openmpi_flexible_path
mirrors/karpathy--llm.c:feature/outliers
mirrors/karpathy--llm.c:feature/rolling_checkpoints
mirrors/karpathy--llm.c:feature/lr_schedulers
mirrors/karpathy--llm.c:feature/nccl-only
mirrors/karpathy--llm.c:feature/hf_eval
mirrors/karpathy--llm.c:feature/permuted_dataloader
mirrors/karpathy--llm.c:feature/delete_backward_act
mirrors/karpathy--llm.c:feature/overlap-backward-reduce
mirrors/karpathy--llm.c:feature/streams
mirrors/karpathy--llm.c:move/matmul
mirrors/karpathy--llm.c:move/fused_classifier
mirrors/karpathy--llm.c:move/attention
mirrors/karpathy--llm.c:move/gelu
mirrors/karpathy--llm.c:move/encoder_kernel
mirrors/karpathy--llm.c:feature/refactor_part2
mirrors/karpathy--llm.c:try/move_cudnn_to_llmc
mirrors/karpathy--llm.c:feature/libdir
mirrors/karpathy--llm.c:feature/pytorch_parity
mirrors/karpathy--llm.c:feature/resume
mirrors/karpathy--llm.c:attempt_fix_zero1
mirrors/karpathy--llm.c:feature/write_checkpoints
mirrors/karpathy--llm.c:feature/weight_decay_2d_only
mirrors/karpathy--llm.c:feature/random_init
mirrors/karpathy--llm.c:feature/hellaswagc
mirrors/karpathy--llm.c:feature/sharded_data
mirrors/karpathy--llm.c:feature/organize_data
mirrors/karpathy--llm.c:feature/hellaswag
mirrors/karpathy--llm.c:feature/recompute
mirrors/karpathy--llm.c:feature/grad_accum3
mirrors/karpathy--llm.c:feature/wikitext103
mirrors/karpathy--llm.c:feature/grad_accum
mirrors/karpathy--llm.c:feature/bias_backward2
mirrors/karpathy--llm.c:feature/fp32_weight_master_copy
mirrors/karpathy--llm.c:feature/cudnn
mirrors/karpathy--llm.c:feature/packed128
mirrors/karpathy--llm.c:separate_tokenizer
mirrors/karpathy--llm.c:ademeure-layernormbf16-3
mirrors/karpathy--llm.c:feature/load_bf16
mirrors/karpathy--llm.c:feature/fix_int_overflow
mirrors/karpathy--llm.c:fix/illegalaccess
mirrors/karpathy--llm.c:feature/cublaslt
..
pull from: mirrors/karpathy--llm.c:main
mirrors/karpathy--llm.c:main
mirrors/karpathy--llm.c:llama3
mirrors/karpathy--llm.c:master
mirrors/karpathy--llm.c:feature/managed2
mirrors/karpathy--llm.c:gordicaleksa-fix_dataloader2
mirrors/karpathy--llm.c:feature/finetune_llama31py
mirrors/karpathy--llm.c:feature/restore_from_master
mirrors/karpathy--llm.c:refactor/code-to-zerocuh
mirrors/karpathy--llm.c:feature/gpt3v1
mirrors/karpathy--llm.c:fix/reshuffle_zero_grad
mirrors/karpathy--llm.c:overfit-batch-actually
mirrors/karpathy--llm.c:feature/openmpi_flexible_path
mirrors/karpathy--llm.c:feature/outliers
mirrors/karpathy--llm.c:feature/rolling_checkpoints
mirrors/karpathy--llm.c:feature/lr_schedulers
mirrors/karpathy--llm.c:feature/nccl-only
mirrors/karpathy--llm.c:feature/hf_eval
mirrors/karpathy--llm.c:feature/permuted_dataloader
mirrors/karpathy--llm.c:feature/delete_backward_act
mirrors/karpathy--llm.c:feature/overlap-backward-reduce
mirrors/karpathy--llm.c:feature/streams
mirrors/karpathy--llm.c:move/matmul
mirrors/karpathy--llm.c:move/fused_classifier
mirrors/karpathy--llm.c:move/attention
mirrors/karpathy--llm.c:move/gelu
mirrors/karpathy--llm.c:move/encoder_kernel
mirrors/karpathy--llm.c:feature/refactor_part2
mirrors/karpathy--llm.c:try/move_cudnn_to_llmc
mirrors/karpathy--llm.c:feature/libdir
mirrors/karpathy--llm.c:feature/pytorch_parity
mirrors/karpathy--llm.c:feature/resume
mirrors/karpathy--llm.c:attempt_fix_zero1
mirrors/karpathy--llm.c:feature/write_checkpoints
mirrors/karpathy--llm.c:feature/weight_decay_2d_only
mirrors/karpathy--llm.c:feature/random_init
mirrors/karpathy--llm.c:feature/hellaswagc
mirrors/karpathy--llm.c:feature/sharded_data
mirrors/karpathy--llm.c:feature/organize_data
mirrors/karpathy--llm.c:feature/hellaswag
mirrors/karpathy--llm.c:feature/recompute
mirrors/karpathy--llm.c:feature/grad_accum3
mirrors/karpathy--llm.c:feature/wikitext103
mirrors/karpathy--llm.c:feature/grad_accum
mirrors/karpathy--llm.c:feature/bias_backward2
mirrors/karpathy--llm.c:feature/fp32_weight_master_copy
mirrors/karpathy--llm.c:feature/cudnn
mirrors/karpathy--llm.c:feature/packed128
mirrors/karpathy--llm.c:separate_tokenizer
mirrors/karpathy--llm.c:ademeure-layernormbf16-3
mirrors/karpathy--llm.c:feature/load_bf16
mirrors/karpathy--llm.c:feature/fix_int_overflow
mirrors/karpathy--llm.c:fix/illegalaccess
mirrors/karpathy--llm.c:feature/cublaslt
These branches are equal.