Update requirements.txt #1245
Open
Update requirements.txt #1245
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The latest version of the Paddleformers library is 0.2.1. If the default download is the latest, the following error will occur. ``` /opt/conda/envs/python35-paddle120-env/lib/python3.10/site-packages/paddle/utils/cpp_extension/extension_utils.py:717: UserWarning: No ccache found. Please be aware that recompiling all source files may be required. You can download and install ccache from: https://github.com/ccache/ccache/blob/master/doc/INSTALL.md warnings.warn(warning_message) None of PyTorch, TensorFlow >= 2.0, or Flax have been found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. /opt/conda/envs/python35-paddle120-env/lib/python3.10/site-packages/_distutils_hack/__init__.py:30: UserWarning: Setuptools is replacing distutils. Support for replacing an already imported distutils is deprecated. In the future, this condition will fail. Register concerns at https://github.com/pypa/setuptools/issues/new?template=distutils-deprecation.yml warnings.warn( W0907 09:45:14.021631 3142 gpu_resources.cc:114] Please NOTE: device: 0, GPU Compute Capability: 8.0, Driver API Version: 12.8, Runtime API Version: 12.6 /opt/conda/envs/python35-paddle120-env/lib/python3.10/site-packages/paddle/utils/cpp_extension/extension_utils.py:717: UserWarning: No ccache found. Please be aware that recompiling all source files may be required. You can download and install ccache from: https://github.com/ccache/ccache/blob/master/doc/INSTALL.md warnings.warn(warning_message) LAUNCH INFO 2025-09-07 09:45:17,460 ----------- Configuration ---------------------- LAUNCH INFO 2025-09-07 09:45:17,460 auto_cluster_config: 0 LAUNCH INFO 2025-09-07 09:45:17,460 auto_parallel_config: None LAUNCH INFO 2025-09-07 09:45:17,460 auto_tuner_json: None LAUNCH INFO 2025-09-07 09:45:17,460 devices: 0 LAUNCH INFO 2025-09-07 09:45:17,460 elastic_level: -1 LAUNCH INFO 2025-09-07 09:45:17,460 elastic_timeout: 30 LAUNCH INFO 2025-09-07 09:45:17,460 enable_gpu_log: True LAUNCH INFO 2025-09-07 09:45:17,460 gloo_port: 6767 LAUNCH INFO 2025-09-07 09:45:17,460 host: None LAUNCH INFO 2025-09-07 09:45:17,460 ips: None LAUNCH INFO 2025-09-07 09:45:17,460 job_id: default LAUNCH INFO 2025-09-07 09:45:17,460 legacy: False LAUNCH INFO 2025-09-07 09:45:17,460 log_dir: erniekit_dist_log LAUNCH INFO 2025-09-07 09:45:17,460 log_level: INFO LAUNCH INFO 2025-09-07 09:45:17,460 log_overwrite: False LAUNCH INFO 2025-09-07 09:45:17,460 master: 127.0.0.1:8080 LAUNCH INFO 2025-09-07 09:45:17,460 max_restart: 3 LAUNCH INFO 2025-09-07 09:45:17,460 nnodes: 1 LAUNCH INFO 2025-09-07 09:45:17,460 nproc_per_node: None LAUNCH INFO 2025-09-07 09:45:17,460 rank: -1 LAUNCH INFO 2025-09-07 09:45:17,460 run_mode: collective LAUNCH INFO 2025-09-07 09:45:17,460 server_num: None LAUNCH INFO 2025-09-07 09:45:17,460 servers: LAUNCH INFO 2025-09-07 09:45:17,460 sort_ip: False LAUNCH INFO 2025-09-07 09:45:17,460 start_port: 6070 LAUNCH INFO 2025-09-07 09:45:17,460 trainer_num: None LAUNCH INFO 2025-09-07 09:45:17,460 trainers: LAUNCH INFO 2025-09-07 09:45:17,461 training_script: /home/aistudio/ERNIE/erniekit/launcher.py LAUNCH INFO 2025-09-07 09:45:17,461 training_script_args: ['train', '/home/aistudio/configs/ERNIE-4.5-0.3B/sft/run_sft_lora_32k.yaml'] LAUNCH INFO 2025-09-07 09:45:17,461 with_gloo: 1 LAUNCH INFO 2025-09-07 09:45:17,461 -------------------------------------------------- LAUNCH INFO 2025-09-07 09:45:17,461 Job: default, mode collective, replicas 1[1:1], elastic False LAUNCH INFO 2025-09-07 09:45:17,462 Run Pod: wsqqoa, replicas 1, status ready LAUNCH INFO 2025-09-07 09:45:17,474 Watching Pod: wsqqoa, replicas 1, status running /opt/conda/envs/python35-paddle120-env/lib/python3.10/site-packages/paddle/utils/cpp_extension/extension_utils.py:717: UserWarning: No ccache found. Please be aware that recompiling all source files may be required. You can download and install ccache from: https://github.com/ccache/ccache/blob/master/doc/INSTALL.md warnings.warn(warning_message) None of PyTorch, TensorFlow >= 2.0, or Flax have been found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. /opt/conda/envs/python35-paddle120-env/lib/python3.10/site-packages/_distutils_hack/__init__.py:30: UserWarning: Setuptools is replacing distutils. Support for replacing an already imported distutils is deprecated. In the future, this condition will fail. Register concerns at https://github.com/pypa/setuptools/issues/new?template=distutils-deprecation.yml warnings.warn( W0907 09:45:21.091935 3501 gpu_resources.cc:114] Please NOTE: device: 0, GPU Compute Capability: 8.0, Driver API Version: 12.8, Runtime API Version: 12.6 [2025-09-07 09:45:21,425] [ INFO] - user has defined resume_from_checkpoint: None [2025-09-07 09:45:21,426] [ INFO] - The default value for the training argument `--report_to` will change in v5 (from all installed integrations to none). In v5, you will need to use `--report_to all` to get the same behavior as now. You should start updating your code and make this info disappear :-). [2025-09-07 09:45:21,426] [ INFO] - reset finetuning arguments global_batch_size to 8 [2025-09-07 09:45:21,426] [ WARNING] - eval_batch_size set to 1 [2025-09-07 09:45:21,457] [ INFO] - Tensor_parallel_degree = 1. Set sequence_parallel to False. [2025-09-07 09:45:21,458] [ DEBUG] - ============================================================ [2025-09-07 09:45:21,458] [ DEBUG] - Model Configuration Arguments [2025-09-07 09:45:21,458] [ DEBUG] - paddle commit id : 0cd8342db88f96ef57398e888b050b5077ae6e73 [2025-09-07 09:45:21,458] [ DEBUG] - paddleformers commit id : c8e46d9f40ac234baac500b0621873f2da7b9d47 [2025-09-07 09:45:21,458] [ DEBUG] - add_tail_layers : False [2025-09-07 09:45:21,458] [ DEBUG] - bos_token_id : 0 [2025-09-07 09:45:21,458] [ DEBUG] - continue_training : True [2025-09-07 09:45:21,458] [ DEBUG] - download_hub : None [2025-09-07 09:45:21,458] [ DEBUG] - eos_token_id : 1 [2025-09-07 09:45:21,459] [ DEBUG] - fine_tuning : LoRA [2025-09-07 09:45:21,459] [ DEBUG] - fuse_gate_detach_matmul : True [2025-09-07 09:45:21,459] [ DEBUG] - fuse_linear : False [2025-09-07 09:45:21,459] [ DEBUG] - fuse_rms_norm : True [2025-09-07 09:45:21,459] [ DEBUG] - fuse_rope : True [2025-09-07 09:45:21,459] [ DEBUG] - fuse_softmax_mask : False [2025-09-07 09:45:21,459] [ DEBUG] - fuse_swiglu : True [2025-09-07 09:45:21,459] [ DEBUG] - lora : True [2025-09-07 09:45:21,459] [ DEBUG] - lora_alpha : -1 [2025-09-07 09:45:21,459] [ DEBUG] - lora_path : None [2025-09-07 09:45:21,459] [ DEBUG] - lora_plus_scale : 1.0 [2025-09-07 09:45:21,459] [ DEBUG] - lora_rank : 32 [2025-09-07 09:45:21,459] [ DEBUG] - loss_subbatch_seqlen : 32768 [2025-09-07 09:45:21,459] [ DEBUG] - max_position_embeddings : 4096 [2025-09-07 09:45:21,459] [ DEBUG] - model_name_or_path : /home/aistudio/model/ERNIE-4.5-0.3B-Paddle [2025-09-07 09:45:21,459] [ DEBUG] - moe_aux_loss_lambda : 1e-05 [2025-09-07 09:45:21,459] [ DEBUG] - moe_gate : top2_fused [2025-09-07 09:45:21,459] [ DEBUG] - moe_group : dummy [2025-09-07 09:45:21,459] [ DEBUG] - moe_group_experts : False [2025-09-07 09:45:21,459] [ DEBUG] - moe_multimodal_dispatch_use_allgather: v2-alltoall-unpad [2025-09-07 09:45:21,459] [ DEBUG] - moe_orthogonal_loss_lambda : 0.0 [2025-09-07 09:45:21,459] [ DEBUG] - moe_use_aux_free : None [2025-09-07 09:45:21,460] [ DEBUG] - moe_use_hard_gate : False [2025-09-07 09:45:21,460] [ DEBUG] - moe_with_send_router_loss : False [2025-09-07 09:45:21,460] [ DEBUG] - moe_z_loss_lambda : 0.0 [2025-09-07 09:45:21,460] [ DEBUG] - no_recompute_layers : None [2025-09-07 09:45:21,460] [ DEBUG] - num_nextn_predict_layers : 0 [2025-09-07 09:45:21,460] [ DEBUG] - offload_recompute_inputs : False [2025-09-07 09:45:21,460] [ DEBUG] - pp_seg_method : layer:Ernie4_5_DecoderLayer|ErnieDecoderLayer|EmptyLayer [2025-09-07 09:45:21,460] [ DEBUG] - recompute_granularity : full [2025-09-07 09:45:21,460] [ DEBUG] - recompute_use_reentrant : True [2025-09-07 09:45:21,460] [ DEBUG] - rope_3d : True [2025-09-07 09:45:21,460] [ DEBUG] - rslora : False [2025-09-07 09:45:21,460] [ DEBUG] - rslora_plus : False [2025-09-07 09:45:21,460] [ DEBUG] - stage : SFT [2025-09-07 09:45:21,460] [ DEBUG] - tensor_parallel_output : True [2025-09-07 09:45:21,460] [ DEBUG] - use_attn_mask_start_row_indices: True [2025-09-07 09:45:21,460] [ DEBUG] - use_flash_attention : True [2025-09-07 09:45:21,460] [ DEBUG] - use_flash_attn_with_mask : True [2025-09-07 09:45:21,460] [ DEBUG] - use_fused_head_and_loss_fn : False [2025-09-07 09:45:21,460] [ DEBUG] - use_mem_eff_attn : True [2025-09-07 09:45:21,460] [ DEBUG] - use_recompute_loss_fn : True [2025-09-07 09:45:21,460] [ DEBUG] - use_recompute_moe : False [2025-09-07 09:45:21,460] [ DEBUG] - use_sparse_flash_attn : True [2025-09-07 09:45:21,460] [ DEBUG] - use_sparse_head_and_loss_fn : True [2025-09-07 09:45:21,461] [ DEBUG] - virtual_pp_degree : 1 [2025-09-07 09:45:21,461] [ DEBUG] - vision_config : VisionArguments(attn_implementation='eager', attn_sep=True, depth=32, embed_dim=1280, hidden_act='quick_gelu', hidden_size=1280, in_channels=3, in_chans=3, mlp_ratio=4, model_type='DFNRope_vision_transformer', num_heads=16, patch_size=14, spatial_merge_size=2, spatial_patch_size=14, tensor_parallel_degree=4, use_recompute=True, vit_num_recompute_layers=10000) [2025-09-07 09:45:21,461] [ DEBUG] - [2025-09-07 09:45:21,461] [ DEBUG] - ============================================================ [2025-09-07 09:45:21,461] [ DEBUG] - Data Configuration Arguments [2025-09-07 09:45:21,461] [ DEBUG] - paddle commit id : 0cd8342db88f96ef57398e888b050b5077ae6e73 [2025-09-07 09:45:21,461] [ DEBUG] - paddleformers commit id : c8e46d9f40ac234baac500b0621873f2da7b9d47 [2025-09-07 09:45:21,461] [ DEBUG] - buffer_size : 500 [2025-09-07 09:45:21,461] [ DEBUG] - dataset_name : KnowledgeBasedSFTReader [2025-09-07 09:45:21,461] [ DEBUG] - dataset_type : iterable [2025-09-07 09:45:21,461] [ DEBUG] - eval_dataset_path : /home/aistudio/datasets/val.jsonl [2025-09-07 09:45:21,461] [ DEBUG] - eval_dataset_prob : 1.0 [2025-09-07 09:45:21,461] [ DEBUG] - eval_dataset_type : erniekit [2025-09-07 09:45:21,461] [ DEBUG] - greedy_intokens : True [2025-09-07 09:45:21,461] [ DEBUG] - in_tokens_batching : True [2025-09-07 09:45:21,461] [ DEBUG] - mask_out_eos_token : True [2025-09-07 09:45:21,461] [ DEBUG] - max_prompt_len : 2048 [2025-09-07 09:45:21,461] [ DEBUG] - max_seq_len : 32768 [2025-09-07 09:45:21,461] [ DEBUG] - num_comparisons : 6 [2025-09-07 09:45:21,461] [ DEBUG] - num_samples_each_epoch : 6000000 [2025-09-07 09:45:21,461] [ DEBUG] - offline_dataset_path : None [2025-09-07 09:45:21,462] [ DEBUG] - random_shuffle : True [2025-09-07 09:45:21,462] [ DEBUG] - text_dataset_path : None [2025-09-07 09:45:21,462] [ DEBUG] - text_dataset_prob : None [2025-09-07 09:45:21,462] [ DEBUG] - train_dataset_path : /home/aistudio/datasets/train.jsonl [2025-09-07 09:45:21,462] [ DEBUG] - train_dataset_prob : 1.0 [2025-09-07 09:45:21,462] [ DEBUG] - train_dataset_type : erniekit [2025-09-07 09:45:21,462] [ DEBUG] - use_cls : True [2025-09-07 09:45:21,462] [ DEBUG] - [2025-09-07 09:45:21,462] [ INFO] - The global seed is set to 23, local seed is set to 24 and random seed is set to 23. [2025-09-07 09:45:21,462] [ WARNING] - Process rank: -1, device: gpu, world_size: 1, distributed training: False, 16-bits training: True [2025-09-07 09:45:21,462] [ INFO] - Start to load model ... [2025-09-07 09:45:21,463] [ INFO] - Using download source: huggingface [2025-09-07 09:45:21,463] [ INFO] - Loading configuration file /home/aistudio/model/ERNIE-4.5-0.3B-Paddle/config.json [2025-09-07 09:45:21,463] [ WARNING] - You are using a model of type ernie4_5 to instantiate a model of type ernie4_5_moe. This is not supported for all configurations of models and can yield errors. [2025-09-07 09:45:21,464] [ INFO] - Loading weights file /home/aistudio/model/ERNIE-4.5-0.3B-Paddle/model.safetensors [2025-09-07 09:45:21,870] [ INFO] - Loaded weights file from disk, setting weights to model. Traceback (most recent call last): File "/home/aistudio/ERNIE/erniekit/launcher.py", line 46, in <module> launch() File "/home/aistudio/ERNIE/erniekit/launcher.py", line 34, in launch run_tuner() File "/home/aistudio/ERNIE/erniekit/train/tuner.py", line 76, in run_tuner _training_function(config={"args": args}) File "/home/aistudio/ERNIE/erniekit/train/tuner.py", line 55, in _training_function run_sft(model_args, data_args, generating_args, finetuning_args) File "/home/aistudio/ERNIE/erniekit/train/sft/workflow.py", line 362, in run_sft model = model_class.from_pretrained( File "/home/aistudio/external-libraries/lib/python3.10/site-packages/paddleformers/transformers/model_utils.py", line 2665, in from_pretrained model = cls(config, *init_args, **model_kwargs) File "/home/aistudio/external-libraries/lib/python3.10/site-packages/paddleformers/transformers/utils.py", line 290, in __impl__ init_func(self, *args, **kwargs) TypeError: Ernie4_5_MoeForCausalLM.__init__() got an unexpected keyword argument 'convert_from_hf' LAUNCH INFO 2025-09-07 09:45:22,480 Pod failed LAUNCH ERROR 2025-09-07 09:45:22,480 Container failed !!! Container rank 0 status failed cmd ['/opt/conda/envs/python35-paddle120-env/bin/python', '-u', '/home/aistudio/ERNIE/erniekit/launcher.py', 'train', '/home/aistudio/configs/ERNIE-4.5-0.3B/sft/run_sft_lora_32k.yaml'] code 1 log erniekit_dist_log/workerlog.0 LAUNCH INFO 2025-09-07 09:45:22,480 ------------------------- ERROR LOG DETAIL ------------------------- m[2025-09-07 09:45:21,461] [ DEBUG] - num_samples_each_epoch : 6000000 [2025-09-07 09:45:21,461] [ DEBUG] - offline_dataset_path : None [2025-09-07 09:45:21,462] [ DEBUG] - random_shuffle : True [2025-09-07 09:45:21,462] [ DEBUG] - text_dataset_path : None [2025-09-07 09:45:21,462] [ DEBUG] - text_dataset_prob : None [2025-09-07 09:45:21,462] [ DEBUG] - train_dataset_path : /home/aistudio/datasets/train.jsonl [2025-09-07 09:45:21,462] [ DEBUG] - train_dataset_prob : 1.0 [2025-09-07 09:45:21,462] [ DEBUG] - train_dataset_type : erniekit [2025-09-07 09:45:21,462] [ DEBUG] - use_cls : True [2025-09-07 09:45:21,462] [ DEBUG] - [2025-09-07 09:45:21,462] [ INFO] - The global seed is set to 23, local seed is set to 24 and random seed is set to 23. [2025-09-07 09:45:21,462] [ WARNING] - Process rank: -1, device: gpu, world_size: 1, distributed training: False, 16-bits training: True [2025-09-07 09:45:21,462] [ INFO] - Start to load model ... [2025-09-07 09:45:21,463] [ INFO] - Using download source: huggingface [2025-09-07 09:45:21,463] [ INFO] - Loading configuration file /home/aistudio/model/ERNIE-4.5-0.3B-Paddle/config.json [2025-09-07 09:45:21,463] [ WARNING] - You are using a model of type ernie4_5 to instantiate a model of type ernie4_5_moe. This is not supported for all configurations of models and can yield errors. [2025-09-07 09:45:21,464] [ INFO] - Loading weights file /home/aistudio/model/ERNIE-4.5-0.3B-Paddle/model.safetensors [2025-09-07 09:45:21,870] [ INFO] - Loaded weights file from disk, setting weights to model. Traceback (most recent call last): File "/home/aistudio/ERNIE/erniekit/launcher.py", line 46, in <module> launch() File "/home/aistudio/ERNIE/erniekit/launcher.py", line 34, in launch run_tuner() File "/home/aistudio/ERNIE/erniekit/train/tuner.py", line 76, in run_tuner _training_function(config={"args": args}) File "/home/aistudio/ERNIE/erniekit/train/tuner.py", line 55, in _training_function run_sft(model_args, data_args, generating_args, finetuning_args) File "/home/aistudio/ERNIE/erniekit/train/sft/workflow.py", line 362, in run_sft model = model_class.from_pretrained( File "/home/aistudio/external-libraries/lib/python3.10/site-packages/paddleformers/transformers/model_utils.py", line 2665, in from_pretrained model = cls(config, *init_args, **model_kwargs) File "/home/aistudio/external-libraries/lib/python3.10/site-packages/paddleformers/transformers/utils.py", line 290, in __impl__ init_func(self, *args, **kwargs) TypeError: Ernie4_5_MoeForCausalLM.__init__() got an unexpected keyword argument 'convert_from_hf' LAUNCH INFO 2025-09-07 09:45:22,480 Exit code 1 ```
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The latest version of the Paddleformers library is 0.2.1. If the default download is the latest, the following error will occur.