chore: import upstream snapshot with attribution

2026-07-13 13:34:58 +08:00
commit a203934033
1368 changed files with 175001 additions and 0 deletions
@@ -0,0 +1,47 @@
+# 2 * 76GiB
+PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
+NPROC_PER_NODE=2 \
+IMAGE_MAX_TOKEN_NUM=1024 \
+VIDEO_MAX_TOKEN_NUM=128 \
+FPS_MAX_FRAMES=16 \
+CUDA_VISIBLE_DEVICES=0,1 \
+megatron sft \
+    --model Qwen/Qwen3-VL-8B-Instruct \
+    --save_safetensors true \
+    --dataset 'AI-ModelScope/LaTeX_OCR:human_handwrite#5000' \
+    --load_from_cache_file true \
+    --tensor_model_parallel_size 2 \
+    --sequence_parallel true \
+    --packing true \
+    --freeze_llm false \
+    --freeze_vit true \
+    --freeze_aligner true \
+    --split_dataset_ratio 0.01 \
+    --micro_batch_size 1 \
+    --global_batch_size 4 \
+    --recompute_granularity full \
+    --recompute_method uniform \
+    --recompute_num_layers 1 \
+    --finetune true \
+    --cross_entropy_loss_fusion true \
+    --lr 1e-5 \
+    --lr_warmup_fraction 0.05 \
+    --min_lr 1e-6 \
+    --num_train_epochs 1 \
+    --output_dir megatron_output/Qwen3-VL-8B-Instruct \
+    --save_steps 200 \
+    --max_length 2048 \
+    --dataloader_num_workers 4 \
+    --no_save_optim true \
+    --no_save_rng true \
+    --dataset_num_proc 8
+
+# PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
+# IMAGE_MAX_TOKEN_NUM=1024 \
+# VIDEO_MAX_TOKEN_NUM=128 \
+# FPS_MAX_FRAMES=16 \
+# CUDA_VISIBLE_DEVICES=0 \
+# swift infer \
+#     --model megatron_output/Qwen3-VL-8B-Instruct/vx-xxx/checkpoint-xxx \
+#     --load_data_args true \
+#     --stream true
@@ -0,0 +1,41 @@
+# 8 * 76GiB, 3s/it
+PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
+CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 \
+NPROC_PER_NODE=8 \
+megatron sft \
+    --model Qwen/Qwen3-30B-A3B-Instruct-2507 \
+    --save_safetensors true \
+    --dataset 'swift/Chinese-Qwen3-235B-Thinking-2507-Distill-data-110k-SFT#20000' \
+    --load_from_cache_file true \
+    --split_dataset_ratio 0.01 \
+    --moe_permute_fusion true \
+    --pipeline_model_parallel_size 2 \
+    --decoder_first_pipeline_num_layers 25 \
+    --tensor_model_parallel_size 4 \
+    --expert_model_parallel_size 4 \
+    --moe_grouped_gemm true \
+    --moe_shared_expert_overlap true \
+    --moe_aux_loss_coeff 1e-6 \
+    --micro_batch_size 1 \
+    --global_batch_size 4 \
+    --recompute_granularity full \
+    --recompute_method uniform \
+    --recompute_num_layers 1 \
+    --num_train_epochs 1 \
+    --finetune true \
+    --cross_entropy_loss_fusion true \
+    --lr 1e-5 \
+    --lr_warmup_fraction 0.05 \
+    --min_lr 1e-6 \
+    --output_dir megatron_output/Qwen3-30B-A3B-Instruct-2507 \
+    --eval_steps 500 \
+    --save_steps 500 \
+    --max_length 8192 \
+    --packing true \
+    --dataloader_num_workers 8 \
+    --dataset_num_proc 8 \
+    --no_save_optim true \
+    --no_save_rng true \
+    --sequence_parallel true \
+    --moe_expert_capacity_factor 2 \
+    --attention_backend flash
@@ -0,0 +1,52 @@
+# 50GiB
+PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
+NPROC_PER_NODE=2 \
+CUDA_VISIBLE_DEVICES=0,1 \
+megatron sft \
+    --model Qwen/Qwen3-30B-A3B-Instruct-2507 \
+    --save_safetensors true \
+    --merge_lora false \
+    --dataset 'swift/Chinese-Qwen3-235B-2507-Distill-data-110k-SFT#2000' \
+              'swift/self-cognition#1000' \
+    --load_from_cache_file true \
+    --tuner_type lora \
+    --lora_rank 8 \
+    --lora_alpha 32 \
+    --target_modules all-linear \
+    --split_dataset_ratio 0.01 \
+    --moe_permute_fusion true \
+    --expert_model_parallel_size 2 \
+    --moe_grouped_gemm true \
+    --moe_shared_expert_overlap true \
+    --moe_aux_loss_coeff 1e-3 \
+    --micro_batch_size 8 \
+    --global_batch_size 16 \
+    --recompute_granularity full \
+    --recompute_method uniform \
+    --recompute_num_layers 1 \
+    --num_train_epochs 1 \
+    --finetune true \
+    --cross_entropy_loss_fusion true \
+    --lr 1e-4 \
+    --lr_warmup_fraction 0.05 \
+    --min_lr 1e-5 \
+    --output_dir megatron_output/Qwen3-30B-A3B-Instruct-2507 \
+    --eval_steps 200 \
+    --save_steps 200 \
+    --max_length 2048 \
+    --dataloader_num_workers 8 \
+    --dataset_num_proc 8 \
+    --no_save_optim true \
+    --no_save_rng true \
+    --sequence_parallel true \
+    --moe_expert_capacity_factor 2 \
+    --attention_backend flash \
+    --model_author swift \
+    --model_name swift-robot
+
+
+# CUDA_VISIBLE_DEVICES=0 \
+# swift infer \
+#     --model Qwen/Qwen3-30B-A3B-Instruct-2507 \
+#     --adapters megatron_output/Qwen3-30B-A3B-Instruct-2507/vx-xxx/checkpoint-xxx \
+#     --stream true
@@ -0,0 +1,50 @@
+# 2 * 50GiB
+PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
+NPROC_PER_NODE=2 \
+CUDA_VISIBLE_DEVICES=0,1 \
+megatron sft \
+    --model Qwen/Qwen3-30B-A3B-Instruct-2507 \
+    --save_safetensors true \
+    --merge_lora false \
+    --dataset 'swift/new_special_tokens' \
+    --load_from_cache_file true \
+    --new_special_tokens 'examples/train/new_special_tokens/tokens.txt' \
+    --tuner_type lora \
+    --lora_rank 8 \
+    --lora_alpha 32 \
+    --target_modules all-linear \
+    --modules_to_save word_embeddings output_layer \
+    --split_dataset_ratio 0.01 \
+    --expert_model_parallel_size 2 \
+    --moe_permute_fusion true \
+    --moe_grouped_gemm true \
+    --moe_shared_expert_overlap true \
+    --moe_aux_loss_coeff 1e-3 \
+    --micro_batch_size 32 \
+    --global_batch_size 64 \
+    --recompute_granularity full \
+    --recompute_method uniform \
+    --recompute_num_layers 1 \
+    --num_train_epochs 5 \
+    --finetune true \
+    --cross_entropy_loss_fusion true \
+    --lr 1e-4 \
+    --lr_warmup_fraction 0.05 \
+    --min_lr 1e-5 \
+    --output_dir megatron_output/Qwen3-30B-A3B-Instruct-2507 \
+    --eval_steps 500 \
+    --save_steps 500 \
+    --max_length 2048 \
+    --dataloader_num_workers 8 \
+    --dataset_num_proc 8 \
+    --no_save_optim true \
+    --no_save_rng true \
+    --moe_expert_capacity_factor 2 \
+    --sequence_parallel true \
+    --attention_backend flash
+
+# CUDA_VISIBLE_DEVICES=0,1 \
+# swift infer \
+#     --adapters megatron_output/Qwen3-30B-A3B-Instruct-2507/vx-xxx/checkpoint-xxx \
+#     --load_data_args true \
+#     --stream true
@@ -0,0 +1,54 @@
+# 2 * 15GiB
+PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
+NPROC_PER_NODE=2 \
+IMAGE_MAX_TOKEN_NUM=1024 \
+VIDEO_MAX_TOKEN_NUM=128 \
+FPS_MAX_FRAMES=16 \
+CUDA_VISIBLE_DEVICES=0,1 \
+megatron sft \
+    --model Qwen/Qwen3-VL-8B-Instruct \
+    --save_safetensors true \
+    --merge_lora false \
+    --dataset 'tany0699/garbage265#20000' \
+    --load_from_cache_file true \
+    --tuner_type lora \
+    --lora_rank 8 \
+    --lora_alpha 32 \
+    --target_modules all-linear \
+    --tensor_model_parallel_size 2 \
+    --sequence_parallel true \
+    --packing true \
+    --freeze_llm false \
+    --freeze_vit true \
+    --freeze_aligner true \
+    --split_dataset_ratio 0.01 \
+    --micro_batch_size 1 \
+    --global_batch_size 4 \
+    --recompute_granularity full \
+    --recompute_method uniform \
+    --recompute_num_layers 1 \
+    --finetune true \
+    --cross_entropy_loss_fusion true \
+    --lr 1e-4 \
+    --lr_warmup_fraction 0.05 \
+    --min_lr 1e-5 \
+    --num_train_epochs 1 \
+    --output_dir megatron_output/Qwen3-VL-8B-Instruct \
+    --save_steps 200 \
+    --max_length 2048 \
+    --dataloader_num_workers 4 \
+    --no_save_optim true \
+    --no_save_rng true \
+    --num_labels 265 \
+    --task_type seq_cls \
+    --dataset_num_proc 8
+
+# PYTORCH_CUDA_ALLOC_CONF='expandable_segments:True' \
+# IMAGE_MAX_TOKEN_NUM=1024 \
+# VIDEO_MAX_TOKEN_NUM=128 \
+# FPS_MAX_FRAMES=16 \
+# CUDA_VISIBLE_DEVICES=0 \
+# swift infer \
+#     --adapters megatron_output/Qwen3-VL-8B-Instruct/vx-xxx/checkpoint-xxx \
+#     --load_data_args true \
+#     --stream true