├── .gitignore ├── LICENSE ├── README.md ├── assets └── 微信_二维码.jpg ├── datasets └── phase_1 │ ├── label2freq_level_1.json │ └── label2freq_level_2.json └── src ├── Megatron-LM ├── .gitignore ├── .gitlab-ci.yml ├── LICENSE ├── MANIFEST.in ├── README.md ├── examples │ ├── evaluate_retriever_nq.sh │ ├── evaluate_zeroshot_gpt.sh │ ├── finetune_mnli_distributed.sh │ ├── finetune_race_distributed.sh │ ├── finetune_retriever_distributed.sh │ ├── generate_text.sh │ ├── merge_mp_bert.sh │ ├── pretrain_bert.sh │ ├── pretrain_bert_distributed.sh │ ├── pretrain_bert_distributed_with_mp.sh │ ├── pretrain_gpt.sh │ ├── pretrain_gpt3_175B.sh │ ├── pretrain_gpt_distributed.sh │ ├── pretrain_gpt_distributed_with_mp.sh │ ├── pretrain_ict.sh │ ├── pretrain_t5.sh │ ├── pretrain_t5_distributed.sh │ ├── pretrain_t5_distributed_with_mp.sh │ ├── run_text_generation_server_345M.sh │ ├── run_text_generation_server_345M_8_tensor_parallel.sh │ └── sc21 │ │ ├── CONFIG.sh │ │ ├── README.md │ │ ├── SBATCH.sh │ │ ├── SRUN.sh │ │ ├── run_figure_11.sh │ │ ├── run_figure_12.sh │ │ ├── run_figure_13.sh │ │ ├── run_figure_14.sh │ │ ├── run_figure_15.sh │ │ ├── run_figure_16.sh │ │ ├── run_figure_17.sh │ │ ├── run_figure_18.sh │ │ └── run_table_1.sh ├── images │ └── cases_april2021.png ├── megatron │ ├── __init__.py │ ├── arguments.py │ ├── checkpointing.py │ ├── data │ │ ├── Makefile │ │ ├── __init__.py │ │ ├── autoaugment.py │ │ ├── bert_dataset.py │ │ ├── biencoder_dataset_utils.py │ │ ├── blendable_dataset.py │ │ ├── data_samplers.py │ │ ├── dataset_utils.py │ │ ├── gpt_dataset.py │ │ ├── helpers.cpp │ │ ├── ict_dataset.py │ │ ├── indexed_dataset.py │ │ ├── orqa_wiki_dataset.py │ │ ├── realm_dataset_utils.py │ │ ├── realm_index.py │ │ ├── t5_dataset.py │ │ ├── test │ │ │ ├── test_indexed_dataset.py │ │ │ └── test_preprocess_data.sh │ │ └── vit_dataset.py │ ├── fp16_deprecated │ │ └── loss_scaler.py │ ├── fused_kernels │ │ ├── __init__.py │ │ ├── compat.h │ │ ├── layer_norm_cuda.cpp │ │ ├── layer_norm_cuda_kernel.cu │ │ ├── scaled_masked_softmax.cpp │ │ ├── scaled_masked_softmax.h │ │ ├── scaled_masked_softmax_cuda.cu │ │ ├── scaled_upper_triang_masked_softmax.cpp │ │ ├── scaled_upper_triang_masked_softmax.h │ │ ├── scaled_upper_triang_masked_softmax_cuda.cu │ │ ├── tests │ │ │ ├── __init__.py │ │ │ └── test_fused_kernels.py │ │ └── type_shim.h │ ├── global_vars.py │ ├── indexer.py │ ├── initialize.py │ ├── learning_rates.py │ ├── memory.py │ ├── microbatches.py │ ├── model │ │ ├── __init__.py │ │ ├── bert_model.py │ │ ├── biencoder_model.py │ │ ├── classification.py │ │ ├── distributed.py │ │ ├── enums.py │ │ ├── fused_bias_gelu.py │ │ ├── fused_layer_norm.py │ │ ├── fused_softmax.py │ │ ├── gpt_model.py │ │ ├── language_model.py │ │ ├── module.py │ │ ├── multiple_choice.py │ │ ├── realm_model.py │ │ ├── t5_model.py │ │ ├── transformer.py │ │ ├── utils.py │ │ └── vit_model.py │ ├── mpu │ │ ├── __init__.py │ │ ├── cross_entropy.py │ │ ├── data.py │ │ ├── initialize.py │ │ ├── layers.py │ │ ├── mappings.py │ │ ├── random.py │ │ ├── tests │ │ │ ├── __init__.py │ │ │ ├── commons.py │ │ │ ├── test_cross_entropy.py │ │ │ ├── test_data.py │ │ │ ├── test_initialize.py │ │ │ ├── test_layers.py │ │ │ └── test_random.py │ │ └── utils.py │ ├── optimizer │ │ ├── __init__.py │ │ ├── clip_grads.py │ │ ├── grad_scaler.py │ │ └── optimizer.py │ ├── p2p_communication.py │ ├── package_info.py │ ├── schedules.py │ ├── text_generation_server.py │ ├── text_generation_utils.py │ ├── tokenizer │ │ ├── __init__.py │ │ ├── bert_tokenization.py │ │ ├── gpt2_tokenization.py │ │ └── tokenizer.py │ ├── training.py │ └── utils.py ├── pretrain_bert.py ├── pretrain_gpt.py ├── pretrain_ict.py ├── pretrain_t5.py ├── pretrain_vit.py ├── scripts_20210904 │ ├── README.md │ ├── preprocess_data_ours.py │ └── pretrain_bert_distributed.sh ├── setup.py ├── tasks │ ├── data_utils.py │ ├── ensemble_classifier.py │ ├── eval_utils.py │ ├── finetune_utils.py │ ├── glue │ │ ├── data.py │ │ ├── finetune.py │ │ ├── mnli.py │ │ └── qqp.py │ ├── main.py │ ├── orqa │ │ ├── README.md │ │ ├── evaluate_orqa.py │ │ ├── evaluate_utils.py │ │ ├── supervised │ │ │ ├── data.py │ │ │ ├── eval_utils.py │ │ │ └── finetune.py │ │ └── unsupervised │ │ │ ├── nq.py │ │ │ ├── qa_utils.py │ │ │ └── tokenizers.py │ ├── race │ │ ├── data.py │ │ └── finetune.py │ ├── vision │ │ ├── classification.py │ │ ├── eval_utils.py │ │ ├── finetune_utils.py │ │ └── main.py │ └── zeroshot_gpt │ │ ├── datasets.py │ │ ├── detokenizer.py │ │ └── evaluate.py ├── tests │ └── test_basic.py └── tools │ ├── generate_samples_gpt.py │ ├── linter.py │ ├── merge_mp_partitions.py │ ├── openwebtext │ ├── README.md │ ├── add_id.py │ ├── blacklist_urls.py │ ├── cleanup_dataset.py │ ├── cleanup_fix_dataset.py │ ├── filter_ngrams.py │ ├── find_duplicates.py │ ├── group_duplicate_url.py │ ├── merge_jsons.py │ └── remove_group_duplicates.py │ ├── preprocess_data.py │ ├── run_text_generation_server.py │ └── text_generation_cli.py ├── SimCSE ├── README.md ├── SentEval │ ├── README.md │ ├── data │ │ └── downstream │ │ │ └── download_dataset.sh │ ├── examples │ │ ├── bow.py │ │ ├── gensen.py │ │ ├── googleuse.py │ │ └── skipthought.py │ ├── senteval │ │ ├── __init__.py │ │ ├── binary.py │ │ ├── probing.py │ │ ├── rank.py │ │ ├── sick.py │ │ ├── snli.py │ │ ├── sst.py │ │ ├── sts.py │ │ ├── tools │ │ │ ├── __init__.py │ │ │ ├── ranking.py │ │ │ ├── relatedness.py │ │ │ └── validation.py │ │ ├── trec.py │ │ └── utils.py │ └── setup.py ├── daguan_task │ ├── README.md │ ├── prepare_nli_datasets.py │ ├── run_sup_example_0905.sh │ ├── train.py │ └── trainers.py ├── data │ ├── download_nli.sh │ └── download_wiki.sh ├── demo │ ├── README.md │ ├── flaskdemo.py │ ├── gradiodemo.py │ ├── run_demo_example.sh │ └── static │ │ └── files │ │ ├── all.js │ │ ├── bootstrap.min.js │ │ ├── plogo.png │ │ └── style.css ├── evaluation.py ├── figure │ ├── demo.gif │ └── model.png ├── run_sup_example.sh ├── run_unsup_example.sh ├── setup.py ├── simcse │ ├── __init__.py │ ├── tool.py │ └── trainers.py ├── simcse_to_huggingface.py └── train.py ├── bert_models ├── model_process │ ├── nezha_convert_tf_checkpoint_to_pytorch.py │ └── replace_bert_model_embeddings.py ├── models │ ├── __init__.py │ ├── classifier.py │ ├── file_utils.py │ ├── modeling_bert.py │ ├── modeling_bert_pabee.py │ └── modeling_nezha.py ├── notes │ ├── README_train_0821_0.md │ ├── README_train_0825_0.md │ ├── README_train_0827_0.md │ └── README_train_0828_0.md ├── training │ ├── at_training.py │ ├── configs.py │ ├── data_loader.py │ ├── dice_loss.py │ ├── focal_loss.py │ ├── main.py │ ├── trainer.py │ └── utils.py └── vocab_process │ ├── dict_vocab2freq_0819.json │ ├── get_vocab_freq_from_corpus.py │ ├── get_vocab_mapping.py │ └── vocab_freq │ └── counts.json ├── classic_models ├── models │ ├── aggregator_layer.py │ ├── classifier.py │ ├── embedding_layer.py │ ├── encoders.py │ └── modeling.py ├── modules │ ├── .gitkeep │ ├── avg_pool.py │ ├── child_dynamic_routing.py │ ├── child_rnns.py │ ├── child_sep_conv.py │ ├── identity_op.py │ ├── max_pool.py │ ├── null_op.py │ ├── positional_embedding.py │ └── self_attn_pool.py ├── notes │ ├── README_train_0817.md │ └── results.md ├── training │ ├── data_loader.py │ ├── focal_loss.py │ ├── main.py │ ├── trainer.py │ └── utils.py ├── utils │ ├── model_utils.py │ └── text_utils.py └── word2vec │ ├── jsonline_to_txt.py │ └── train_word2vec.py ├── data_proc ├── label_vocab_process.py ├── sample_length_stats.py └── split_datasets.py ├── official_baseline ├── baseline.ipynb └── baseline.py └── 代码基础知识 ├── gather_官方文档.PNG ├── gather操作.ipynb ├── ntxent_loss解析.ipynb └── supcontrast_loss解析.ipynb /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/.gitignore -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/LICENSE -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/README.md -------------------------------------------------------------------------------- /assets/微信_二维码.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/assets/微信_二维码.jpg -------------------------------------------------------------------------------- /datasets/phase_1/label2freq_level_1.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/datasets/phase_1/label2freq_level_1.json -------------------------------------------------------------------------------- /datasets/phase_1/label2freq_level_2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/datasets/phase_1/label2freq_level_2.json -------------------------------------------------------------------------------- /src/Megatron-LM/.gitignore: -------------------------------------------------------------------------------- 1 | __pycache__ 2 | 3 | # Distribution / packaging 4 | build/ 5 | dist/ 6 | *.egg-info/ -------------------------------------------------------------------------------- /src/Megatron-LM/.gitlab-ci.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/.gitlab-ci.yml -------------------------------------------------------------------------------- /src/Megatron-LM/LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/LICENSE -------------------------------------------------------------------------------- /src/Megatron-LM/MANIFEST.in: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/MANIFEST.in -------------------------------------------------------------------------------- /src/Megatron-LM/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/README.md -------------------------------------------------------------------------------- /src/Megatron-LM/examples/evaluate_retriever_nq.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/evaluate_retriever_nq.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/evaluate_zeroshot_gpt.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/evaluate_zeroshot_gpt.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/finetune_mnli_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/finetune_mnli_distributed.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/finetune_race_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/finetune_race_distributed.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/finetune_retriever_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/finetune_retriever_distributed.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/generate_text.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/generate_text.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/merge_mp_bert.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/merge_mp_bert.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_bert.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_bert.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_bert_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_bert_distributed.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_bert_distributed_with_mp.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_bert_distributed_with_mp.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_gpt.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_gpt.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_gpt3_175B.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_gpt3_175B.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_gpt_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_gpt_distributed.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_gpt_distributed_with_mp.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_gpt_distributed_with_mp.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_ict.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_ict.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_t5.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_t5.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_t5_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_t5_distributed.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/pretrain_t5_distributed_with_mp.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/pretrain_t5_distributed_with_mp.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/run_text_generation_server_345M.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/run_text_generation_server_345M.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/run_text_generation_server_345M_8_tensor_parallel.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/run_text_generation_server_345M_8_tensor_parallel.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/CONFIG.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/CONFIG.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/README.md -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/SBATCH.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/SBATCH.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/SRUN.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/SRUN.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_11.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_11.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_12.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_12.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_13.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_13.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_14.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_14.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_15.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_15.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_16.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_16.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_17.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_17.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_figure_18.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_figure_18.sh -------------------------------------------------------------------------------- /src/Megatron-LM/examples/sc21/run_table_1.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/examples/sc21/run_table_1.sh -------------------------------------------------------------------------------- /src/Megatron-LM/images/cases_april2021.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/images/cases_april2021.png -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/__init__.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/arguments.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/arguments.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/checkpointing.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/checkpointing.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/Makefile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/Makefile -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/__init__.py: -------------------------------------------------------------------------------- 1 | from . import indexed_dataset 2 | -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/autoaugment.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/autoaugment.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/bert_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/bert_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/biencoder_dataset_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/biencoder_dataset_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/blendable_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/blendable_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/data_samplers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/data_samplers.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/dataset_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/dataset_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/gpt_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/gpt_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/helpers.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/helpers.cpp -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/ict_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/ict_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/indexed_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/indexed_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/orqa_wiki_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/orqa_wiki_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/realm_dataset_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/realm_dataset_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/realm_index.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/realm_index.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/t5_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/t5_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/test/test_indexed_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/test/test_indexed_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/test/test_preprocess_data.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/test/test_preprocess_data.sh -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/data/vit_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/data/vit_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fp16_deprecated/loss_scaler.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fp16_deprecated/loss_scaler.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/__init__.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/compat.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/compat.h -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/layer_norm_cuda.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/layer_norm_cuda.cpp -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/layer_norm_cuda_kernel.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/layer_norm_cuda_kernel.cu -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/scaled_masked_softmax.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/scaled_masked_softmax.cpp -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/scaled_masked_softmax.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/scaled_masked_softmax.h -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/scaled_masked_softmax_cuda.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/scaled_masked_softmax_cuda.cu -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/scaled_upper_triang_masked_softmax.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/scaled_upper_triang_masked_softmax.cpp -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/scaled_upper_triang_masked_softmax.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/scaled_upper_triang_masked_softmax.h -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/scaled_upper_triang_masked_softmax_cuda.cu: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/scaled_upper_triang_masked_softmax_cuda.cu -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/tests/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/tests/test_fused_kernels.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/tests/test_fused_kernels.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/fused_kernels/type_shim.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/fused_kernels/type_shim.h -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/global_vars.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/global_vars.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/indexer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/indexer.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/initialize.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/initialize.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/learning_rates.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/learning_rates.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/memory.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/microbatches.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/microbatches.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/__init__.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/bert_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/bert_model.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/biencoder_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/biencoder_model.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/classification.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/classification.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/distributed.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/distributed.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/enums.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/enums.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/fused_bias_gelu.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/fused_bias_gelu.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/fused_layer_norm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/fused_layer_norm.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/fused_softmax.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/fused_softmax.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/gpt_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/gpt_model.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/language_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/language_model.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/module.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/module.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/multiple_choice.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/multiple_choice.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/realm_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/realm_model.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/t5_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/t5_model.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/transformer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/transformer.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/model/vit_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/model/vit_model.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/__init__.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/cross_entropy.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/cross_entropy.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/data.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/initialize.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/initialize.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/layers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/layers.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/mappings.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/mappings.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/random.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/random.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/tests/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/tests/commons.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/tests/commons.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/tests/test_cross_entropy.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/tests/test_cross_entropy.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/tests/test_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/tests/test_data.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/tests/test_initialize.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/tests/test_initialize.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/tests/test_layers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/tests/test_layers.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/tests/test_random.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/tests/test_random.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/mpu/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/mpu/utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/optimizer/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/optimizer/__init__.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/optimizer/clip_grads.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/optimizer/clip_grads.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/optimizer/grad_scaler.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/optimizer/grad_scaler.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/optimizer/optimizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/optimizer/optimizer.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/p2p_communication.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/p2p_communication.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/package_info.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/package_info.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/schedules.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/schedules.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/text_generation_server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/text_generation_server.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/text_generation_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/text_generation_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/tokenizer/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/tokenizer/__init__.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/tokenizer/bert_tokenization.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/tokenizer/bert_tokenization.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/tokenizer/gpt2_tokenization.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/tokenizer/gpt2_tokenization.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/tokenizer/tokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/tokenizer/tokenizer.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/training.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/training.py -------------------------------------------------------------------------------- /src/Megatron-LM/megatron/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/megatron/utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/pretrain_bert.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/pretrain_bert.py -------------------------------------------------------------------------------- /src/Megatron-LM/pretrain_gpt.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/pretrain_gpt.py -------------------------------------------------------------------------------- /src/Megatron-LM/pretrain_ict.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/pretrain_ict.py -------------------------------------------------------------------------------- /src/Megatron-LM/pretrain_t5.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/pretrain_t5.py -------------------------------------------------------------------------------- /src/Megatron-LM/pretrain_vit.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/pretrain_vit.py -------------------------------------------------------------------------------- /src/Megatron-LM/scripts_20210904/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/scripts_20210904/README.md -------------------------------------------------------------------------------- /src/Megatron-LM/scripts_20210904/preprocess_data_ours.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/scripts_20210904/preprocess_data_ours.py -------------------------------------------------------------------------------- /src/Megatron-LM/scripts_20210904/pretrain_bert_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/scripts_20210904/pretrain_bert_distributed.sh -------------------------------------------------------------------------------- /src/Megatron-LM/setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/setup.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/data_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/data_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/ensemble_classifier.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/ensemble_classifier.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/eval_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/eval_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/finetune_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/finetune_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/glue/data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/glue/data.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/glue/finetune.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/glue/finetune.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/glue/mnli.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/glue/mnli.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/glue/qqp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/glue/qqp.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/main.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/README.md -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/evaluate_orqa.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/evaluate_orqa.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/evaluate_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/evaluate_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/supervised/data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/supervised/data.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/supervised/eval_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/supervised/eval_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/supervised/finetune.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/supervised/finetune.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/unsupervised/nq.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/unsupervised/nq.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/unsupervised/qa_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/unsupervised/qa_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/orqa/unsupervised/tokenizers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/orqa/unsupervised/tokenizers.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/race/data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/race/data.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/race/finetune.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/race/finetune.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/vision/classification.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/vision/classification.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/vision/eval_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/vision/eval_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/vision/finetune_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/vision/finetune_utils.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/vision/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/vision/main.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/zeroshot_gpt/datasets.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/zeroshot_gpt/datasets.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/zeroshot_gpt/detokenizer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/zeroshot_gpt/detokenizer.py -------------------------------------------------------------------------------- /src/Megatron-LM/tasks/zeroshot_gpt/evaluate.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tasks/zeroshot_gpt/evaluate.py -------------------------------------------------------------------------------- /src/Megatron-LM/tests/test_basic.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tests/test_basic.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/generate_samples_gpt.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/generate_samples_gpt.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/linter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/linter.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/merge_mp_partitions.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/merge_mp_partitions.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/README.md -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/add_id.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/add_id.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/blacklist_urls.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/blacklist_urls.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/cleanup_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/cleanup_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/cleanup_fix_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/cleanup_fix_dataset.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/filter_ngrams.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/filter_ngrams.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/find_duplicates.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/find_duplicates.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/group_duplicate_url.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/group_duplicate_url.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/merge_jsons.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/merge_jsons.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/openwebtext/remove_group_duplicates.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/openwebtext/remove_group_duplicates.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/preprocess_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/preprocess_data.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/run_text_generation_server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/run_text_generation_server.py -------------------------------------------------------------------------------- /src/Megatron-LM/tools/text_generation_cli.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/Megatron-LM/tools/text_generation_cli.py -------------------------------------------------------------------------------- /src/SimCSE/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/README.md -------------------------------------------------------------------------------- /src/SimCSE/SentEval/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/README.md -------------------------------------------------------------------------------- /src/SimCSE/SentEval/data/downstream/download_dataset.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/data/downstream/download_dataset.sh -------------------------------------------------------------------------------- /src/SimCSE/SentEval/examples/bow.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/examples/bow.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/examples/gensen.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/examples/gensen.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/examples/googleuse.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/examples/googleuse.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/examples/skipthought.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/examples/skipthought.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/__init__.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/binary.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/binary.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/probing.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/probing.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/rank.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/rank.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/sick.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/sick.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/snli.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/snli.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/sst.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/sst.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/sts.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/sts.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/tools/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/tools/ranking.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/tools/ranking.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/tools/relatedness.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/tools/relatedness.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/tools/validation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/tools/validation.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/trec.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/trec.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/senteval/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/senteval/utils.py -------------------------------------------------------------------------------- /src/SimCSE/SentEval/setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/SentEval/setup.py -------------------------------------------------------------------------------- /src/SimCSE/daguan_task/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/daguan_task/README.md -------------------------------------------------------------------------------- /src/SimCSE/daguan_task/prepare_nli_datasets.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/daguan_task/prepare_nli_datasets.py -------------------------------------------------------------------------------- /src/SimCSE/daguan_task/run_sup_example_0905.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/daguan_task/run_sup_example_0905.sh -------------------------------------------------------------------------------- /src/SimCSE/daguan_task/train.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/daguan_task/train.py -------------------------------------------------------------------------------- /src/SimCSE/daguan_task/trainers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/daguan_task/trainers.py -------------------------------------------------------------------------------- /src/SimCSE/data/download_nli.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/data/download_nli.sh -------------------------------------------------------------------------------- /src/SimCSE/data/download_wiki.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/data/download_wiki.sh -------------------------------------------------------------------------------- /src/SimCSE/demo/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/README.md -------------------------------------------------------------------------------- /src/SimCSE/demo/flaskdemo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/flaskdemo.py -------------------------------------------------------------------------------- /src/SimCSE/demo/gradiodemo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/gradiodemo.py -------------------------------------------------------------------------------- /src/SimCSE/demo/run_demo_example.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/run_demo_example.sh -------------------------------------------------------------------------------- /src/SimCSE/demo/static/files/all.js: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/static/files/all.js -------------------------------------------------------------------------------- /src/SimCSE/demo/static/files/bootstrap.min.js: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/static/files/bootstrap.min.js -------------------------------------------------------------------------------- /src/SimCSE/demo/static/files/plogo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/static/files/plogo.png -------------------------------------------------------------------------------- /src/SimCSE/demo/static/files/style.css: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/demo/static/files/style.css -------------------------------------------------------------------------------- /src/SimCSE/evaluation.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/evaluation.py -------------------------------------------------------------------------------- /src/SimCSE/figure/demo.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/figure/demo.gif -------------------------------------------------------------------------------- /src/SimCSE/figure/model.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/figure/model.png -------------------------------------------------------------------------------- /src/SimCSE/run_sup_example.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/run_sup_example.sh -------------------------------------------------------------------------------- /src/SimCSE/run_unsup_example.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/run_unsup_example.sh -------------------------------------------------------------------------------- /src/SimCSE/setup.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/setup.py -------------------------------------------------------------------------------- /src/SimCSE/simcse/__init__.py: -------------------------------------------------------------------------------- 1 | from .tool import SimCSE 2 | -------------------------------------------------------------------------------- /src/SimCSE/simcse/tool.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/simcse/tool.py -------------------------------------------------------------------------------- /src/SimCSE/simcse/trainers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/simcse/trainers.py -------------------------------------------------------------------------------- /src/SimCSE/simcse_to_huggingface.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/simcse_to_huggingface.py -------------------------------------------------------------------------------- /src/SimCSE/train.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/SimCSE/train.py -------------------------------------------------------------------------------- /src/bert_models/model_process/nezha_convert_tf_checkpoint_to_pytorch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/model_process/nezha_convert_tf_checkpoint_to_pytorch.py -------------------------------------------------------------------------------- /src/bert_models/model_process/replace_bert_model_embeddings.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/model_process/replace_bert_model_embeddings.py -------------------------------------------------------------------------------- /src/bert_models/models/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/models/__init__.py -------------------------------------------------------------------------------- /src/bert_models/models/classifier.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/models/classifier.py -------------------------------------------------------------------------------- /src/bert_models/models/file_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/models/file_utils.py -------------------------------------------------------------------------------- /src/bert_models/models/modeling_bert.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/models/modeling_bert.py -------------------------------------------------------------------------------- /src/bert_models/models/modeling_bert_pabee.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/models/modeling_bert_pabee.py -------------------------------------------------------------------------------- /src/bert_models/models/modeling_nezha.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/models/modeling_nezha.py -------------------------------------------------------------------------------- /src/bert_models/notes/README_train_0821_0.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/notes/README_train_0821_0.md -------------------------------------------------------------------------------- /src/bert_models/notes/README_train_0825_0.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/notes/README_train_0825_0.md -------------------------------------------------------------------------------- /src/bert_models/notes/README_train_0827_0.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/notes/README_train_0827_0.md -------------------------------------------------------------------------------- /src/bert_models/notes/README_train_0828_0.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/notes/README_train_0828_0.md -------------------------------------------------------------------------------- /src/bert_models/training/at_training.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/at_training.py -------------------------------------------------------------------------------- /src/bert_models/training/configs.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/configs.py -------------------------------------------------------------------------------- /src/bert_models/training/data_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/data_loader.py -------------------------------------------------------------------------------- /src/bert_models/training/dice_loss.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/dice_loss.py -------------------------------------------------------------------------------- /src/bert_models/training/focal_loss.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/focal_loss.py -------------------------------------------------------------------------------- /src/bert_models/training/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/main.py -------------------------------------------------------------------------------- /src/bert_models/training/trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/trainer.py -------------------------------------------------------------------------------- /src/bert_models/training/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/training/utils.py -------------------------------------------------------------------------------- /src/bert_models/vocab_process/dict_vocab2freq_0819.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/vocab_process/dict_vocab2freq_0819.json -------------------------------------------------------------------------------- /src/bert_models/vocab_process/get_vocab_freq_from_corpus.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/vocab_process/get_vocab_freq_from_corpus.py -------------------------------------------------------------------------------- /src/bert_models/vocab_process/get_vocab_mapping.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/vocab_process/get_vocab_mapping.py -------------------------------------------------------------------------------- /src/bert_models/vocab_process/vocab_freq/counts.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/bert_models/vocab_process/vocab_freq/counts.json -------------------------------------------------------------------------------- /src/classic_models/models/aggregator_layer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/models/aggregator_layer.py -------------------------------------------------------------------------------- /src/classic_models/models/classifier.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/models/classifier.py -------------------------------------------------------------------------------- /src/classic_models/models/embedding_layer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/models/embedding_layer.py -------------------------------------------------------------------------------- /src/classic_models/models/encoders.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/models/encoders.py -------------------------------------------------------------------------------- /src/classic_models/models/modeling.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/models/modeling.py -------------------------------------------------------------------------------- /src/classic_models/modules/.gitkeep: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/classic_models/modules/avg_pool.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/avg_pool.py -------------------------------------------------------------------------------- /src/classic_models/modules/child_dynamic_routing.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/child_dynamic_routing.py -------------------------------------------------------------------------------- /src/classic_models/modules/child_rnns.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/child_rnns.py -------------------------------------------------------------------------------- /src/classic_models/modules/child_sep_conv.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/child_sep_conv.py -------------------------------------------------------------------------------- /src/classic_models/modules/identity_op.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/identity_op.py -------------------------------------------------------------------------------- /src/classic_models/modules/max_pool.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/max_pool.py -------------------------------------------------------------------------------- /src/classic_models/modules/null_op.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/null_op.py -------------------------------------------------------------------------------- /src/classic_models/modules/positional_embedding.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/positional_embedding.py -------------------------------------------------------------------------------- /src/classic_models/modules/self_attn_pool.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/modules/self_attn_pool.py -------------------------------------------------------------------------------- /src/classic_models/notes/README_train_0817.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/notes/README_train_0817.md -------------------------------------------------------------------------------- /src/classic_models/notes/results.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/notes/results.md -------------------------------------------------------------------------------- /src/classic_models/training/data_loader.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/training/data_loader.py -------------------------------------------------------------------------------- /src/classic_models/training/focal_loss.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/training/focal_loss.py -------------------------------------------------------------------------------- /src/classic_models/training/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/training/main.py -------------------------------------------------------------------------------- /src/classic_models/training/trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/training/trainer.py -------------------------------------------------------------------------------- /src/classic_models/training/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/training/utils.py -------------------------------------------------------------------------------- /src/classic_models/utils/model_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/utils/model_utils.py -------------------------------------------------------------------------------- /src/classic_models/utils/text_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/utils/text_utils.py -------------------------------------------------------------------------------- /src/classic_models/word2vec/jsonline_to_txt.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/word2vec/jsonline_to_txt.py -------------------------------------------------------------------------------- /src/classic_models/word2vec/train_word2vec.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/classic_models/word2vec/train_word2vec.py -------------------------------------------------------------------------------- /src/data_proc/label_vocab_process.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/data_proc/label_vocab_process.py -------------------------------------------------------------------------------- /src/data_proc/sample_length_stats.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/data_proc/sample_length_stats.py -------------------------------------------------------------------------------- /src/data_proc/split_datasets.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/data_proc/split_datasets.py -------------------------------------------------------------------------------- /src/official_baseline/baseline.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/official_baseline/baseline.ipynb -------------------------------------------------------------------------------- /src/official_baseline/baseline.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/official_baseline/baseline.py -------------------------------------------------------------------------------- /src/代码基础知识/gather_官方文档.PNG: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/代码基础知识/gather_官方文档.PNG -------------------------------------------------------------------------------- /src/代码基础知识/gather操作.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/代码基础知识/gather操作.ipynb -------------------------------------------------------------------------------- /src/代码基础知识/ntxent_loss解析.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/代码基础知识/ntxent_loss解析.ipynb -------------------------------------------------------------------------------- /src/代码基础知识/supcontrast_loss解析.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/michael-wzhu/daguan_competition_2021_codes/HEAD/src/代码基础知识/supcontrast_loss解析.ipynb --------------------------------------------------------------------------------