This repository has been archived by the owner on Nov 22, 2022. It is now read-only.
dist training is not work from user defined task #1217
Comments
|
@hudeven @chenyangyu1988 da lao can you guys help me ? |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
I changed the distributed_world_size to 1, the training is running properly. However, it failed with 2.
Below is my config.
{ "distributed_world_size": 2, "include_dirs": [ "transformer-crf" ], "task": { "MyTaggingTask": { "data": { "Data": { "source": { "UDPOSDataSource": { "column_mapping": {"text": "doc", "slots": "tags"}, "udpos_root": "/home/luning/dev/projects/transformer-crf/.data/udpos/en-ud-v2", "fields": ["text", "slots"] } }, "batcher": { "PoolingBatcher": { "train_batch_size": 512, "eval_batch_size": 512, "test_batch_size": 512, "pool_num_batches": 100, "num_shuffled_pools": 4 } } } }, "model": { "decoder": { "activation": "relu", "dropout": 0.0, "freeze": false, "hidden_dims": [], "layer_norm": false, "load_path": null, "out_dim": null, "save_path": null, "shared_module_key": null }, "embedding": { "cpu_only": false, "delimiter": " ", "embed_dim": 512, "embedding_init_range": null, "embedding_init_strategy": "random", "export_input_names": [ "tokens_vals" ], "freeze": false, "load_path": null, "lowercase_tokens": true, "min_freq": 1, "mlp_layer_dims": [], "padding_idx": null, "pretrained_embeddings_path": "", "save_path": null, "shared_module_key": null, "skip_header": true, "vocab_file": "", "vocab_from_all_data": false, "vocab_from_pretrained_embeddings": false, "vocab_from_train_data": true, "vocab_size": 0 }, "inputs": { "slots": { "add_bos_token": false, "add_eos_token": false, "column": "tags", "max_seq_len": null, "use_eos_token_for_bos": false, "tokenizer": { "DoNothingTokenizer": { "do_nothing": "" } }, "vocab": { "build_from_data": true, "size_from_data": 0, "vocab_files": [ ] }, "vocab_file_delimiter": " " }, "tokens": { "add_bos_token": false, "add_eos_token": false, "column": "doc", "max_seq_len": null, "use_eos_token_for_bos": false, "tokenizer": { "DoNothingTokenizer": { "do_nothing": "" } }, "vocab": { "build_from_data": true, "size_from_data": 0, "vocab_files": [ ] }, "vocab_file_delimiter": " " } }, "output_layer": { "CRFOutputLayer": { "freeze": false, "load_path": null, "save_path": null, "shared_module_key": null } }, "representation": { "bidirectional": true, "dropout": 0.9, "freeze": false, "load_path": null, "lstm_dim": 512, "num_layers": 1, "pack_sequence": true, "save_path": null, "shared_module_key": null } }, "trainer": { "TaskTrainer": { "do_eval": true, "early_stop_after": 0, "epochs": 10000, "fp16_args": { "FP16OptimizerFairseq": { "init_loss_scale": 128, "min_loss_scale": 0.0001, "scale_tolerance": 0.0, "scale_window": null, "threshold_loss_scale": null } }, "max_clip_norm": null, "num_accumulated_batches": 1, "num_batches_per_epoch": null, "num_samples_to_log_progress": 1000, "optimizer": { "Adam": { "eps": 1e-08, "lr": 0.001, "weight_decay": 1e-05 } }, "report_train_metrics": true, "scheduler": null, "sparsifier": null, "target_time_limit_seconds": null } } } }, "use_cuda_if_available": true, "use_tensorboard": true, "version": 18 }Errors
How to run
https://github.com/jiangxiluning/pytext_demo
this is my code repo. You can run with below commands
The text was updated successfully, but these errors were encountered: