[BUG] When trying to convert llama2-7b model from HF format to megatron format

[Sun2018421]

Describe the bug
A clear and concise description of what the bug is.

The error is

Traceback (most recent call last):
  File "/usr/lib/python3.12/multiprocessing/process.py", line 314, in _bootstrap
    self.run()
  File "/usr/lib/python3.12/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/workdir/Megatron-LM/tools/checkpoint/saver_megatron.py", line 168, in save_checkpoint
    validate_args(margs)
  File "/workdir/Megatron-LM/megatron/training/arguments.py", line 405, in validate_args
    args.main_grads_dtype = dtype_map[args.main_grads_dtype]
                            ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^
KeyError: torch.float32

To Reproduce
Steps to reproduce the behavior. The easier it is to reproduce the faster it will get maintainer attention.

1. I get the llama2-7b model from https://huggingface.co/meta-llama/Llama-2-7b-hf/tree/main
2. Try the script as follow

#!/bin/bash
TP=1
PP=2
HF_FORMAT_DIR=/workdir/models/llama2-7b
MEGATRON_FORMAT_DIR=/workdir/models/MModels
TOKENIZER_MODEL=${HF_FORMAT_DIR}
python tools/checkpoint/convert.py 
        --bf16 
        --model-type GPT 
        --loader llama_mistral 
        --saver megatron 
        --target-tensor-parallel-size ${TP} 
        --checkpoint-type hf 
        --load-dir ${HF_FORMAT_DIR} 
        --model-size llama2-7B 
        --save-dir ${MEGATRON_FORMAT_DIR} 
        --tokenizer-model ${TOKENIZER_MODEL} 

Stack trace/logs

Loaded loader_llama_mistral as the loader.
Loaded saver_megatron as the saver.
Starting saver...
Starting loader...
<frozen importlib._bootstrap>:488: DeprecationWarning: Type google._upb._message.MessageMapContainer uses PyType_Spec with a metaclass that has custom tp_new. This is deprecated and will no longer be allowed in Python 3.14.
<frozen importlib._bootstrap>:488: DeprecationWarning: Type google._upb._message.ScalarMapContainer uses PyType_Spec with a metaclass that has custom tp_new. This is deprecated and will no longer be allowed in Python 3.14.
/usr/local/lib/python3.12/dist-packages/google/protobuf/internal/well_known_types.py:93: DeprecationWarning: datetime.datetime.utcfromtimestamp() is deprecated and scheduled for removal in a future version. Use timezone-aware objects to represent datetimes in UTC: datetime.datetime.fromtimestamp(timestamp, datetime.UTC).
  _EPOCH_DATETIME_NAIVE = datetime.datetime.utcfromtimestamp(0)
<frozen importlib._bootstrap>:488: DeprecationWarning: Type google._upb._message.MessageMapContainer uses PyType_Spec with a metaclass that has custom tp_new. This is deprecated and will no longer be allowed in Python 3.14.
<frozen importlib._bootstrap>:488: DeprecationWarning: Type google._upb._message.ScalarMapContainer uses PyType_Spec with a metaclass that has custom tp_new. This is deprecated and will no longer be allowed in Python 3.14.
/usr/local/lib/python3.12/dist-packages/google/protobuf/internal/well_known_types.py:93: DeprecationWarning: datetime.datetime.utcfromtimestamp() is deprecated and scheduled for removal in a future version. Use timezone-aware objects to represent datetimes in UTC: datetime.datetime.fromtimestamp(timestamp, datetime.UTC).
  _EPOCH_DATETIME_NAIVE = datetime.datetime.utcfromtimestamp(0)
using world size: 1, data-parallel size: 1, context-parallel size: 1, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 1, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0
using torch.float32 for parameters ...
max_p_e is 4096 and seq_length is 4096
------------------------ arguments ------------------------
  accumulate_allreduce_grads_in_fp32 .............. False
  adam_beta1 ...................................... 0.9
  adam_beta2 ...................................... 0.999
  adam_eps ........................................ 1e-08
  add_bias_linear ................................. False
  add_position_embedding .......................... True
  add_qkv_bias .................................... False
  adlr_autoresume ................................. False
  adlr_autoresume_interval ........................ 1000
  align_grad_reduce ............................... True
  align_param_gather .............................. False
  app_tag_run_name ................................ None
  app_tag_run_version ............................. 0.0.0
  apply_layernorm_1p .............................. False
  apply_query_key_layer_scaling ................... False
  apply_residual_connection_post_layernorm ........ False
  apply_rope_fusion ............................... True
  async_save ...................................... None
  async_tensor_model_parallel_allreduce ........... False
  attention_backend ............................... AttnBackend.auto
  attention_dropout ............................... 0.1
  attention_softmax_in_fp32 ....................... False
  auto_detect_ckpt_format ......................... False
  barrier_with_L1_time ............................ True
  bert_binary_head ................................ True
  bert_embedder_type .............................. megatron
  bert_load ....................................... None
  bf16 ............................................ False
  bias_dropout_fusion ............................. False
  bias_gelu_fusion ................................ False
  bias_swiglu_fusion .............................. True
  biencoder_projection_dim ........................ 0
  biencoder_shared_query_context_model ............ False
  block_data_path ................................. None
  calculate_per_token_loss ........................ False
  check_for_nan_in_loss_and_grad .................. True
  check_for_spiky_loss ............................ False
  check_weight_hash_across_dp_replicas_interval ... None
  ckpt_assume_constant_structure .................. False
  ckpt_convert_format ............................. None
  ckpt_convert_save ............................... None
  ckpt_convert_update_legacy_dist_opt_format ...... False
  ckpt_format ..................................... torch_dist
  ckpt_fully_parallel_load ........................ False
  ckpt_fully_parallel_save ........................ True
  ckpt_fully_parallel_save_deprecated ............. False
  ckpt_step ....................................... None
  classes_fraction ................................ 1.0
  clip_grad ....................................... 1.0
  clone_scatter_output_in_embedding ............... True
  config_logger_dir ............................... 
  consumed_train_samples .......................... 0
  consumed_valid_samples .......................... 0
  context_parallel_size ........................... 1
  cp_comm_type .................................... ['p2p']
  create_attention_mask_in_dataloader ............. True
  cross_entropy_loss_fusion ....................... False
  data_args_path .................................. None
  data_cache_path ................................. None
  data_parallel_random_init ....................... False
  data_parallel_size .............................. 1
  data_path ....................................... None
  data_per_class_fraction ......................... 1.0
  data_sharding ................................... True
  dataloader_type ................................. single
  ddp_average_in_collective ....................... False
  ddp_bucket_size ................................. None
  decoder_first_pipeline_num_layers ............... None
  decoder_last_pipeline_num_layers ................ None
  decoder_num_layers .............................. None
  decoder_seq_length .............................. None
  decoupled_lr .................................... None
  decoupled_min_lr ................................ None
  decrease_batch_size_if_needed ................... False
  defer_embedding_wgrad_compute ................... False
  deprecated_use_mcore_models ..................... False
  deterministic_mode .............................. False
  dino_bottleneck_size ............................ 256
  dino_freeze_last_layer .......................... 1
  dino_head_hidden_size ........................... 2048
  dino_local_crops_number ......................... 10
  dino_local_img_size ............................. 96
  dino_norm_last_layer ............................ False
  dino_teacher_temp ............................... 0.07
  dino_warmup_teacher_temp ........................ 0.04
  dino_warmup_teacher_temp_epochs ................. 30
  disable_straggler_on_startup .................... False
  dist_ckpt_format_deprecated ..................... None
  dist_ckpt_strictness ............................ assume_ok_unexpected
  distribute_saved_activations .................... False
  distributed_backend ............................. nccl
  distributed_timeout_minutes ..................... 10
  embedding_path .................................. None
  empty_unused_memory_level ....................... 0
  enable_ft_package ............................... False
  enable_one_logger ............................... False
  encoder_num_layers .............................. 32
  encoder_pipeline_model_parallel_size ............ 0
  encoder_seq_length .............................. 4096
  encoder_tensor_model_parallel_size .............. 0
  end_weight_decay ................................ 0.01
  eod_mask_loss ................................... False
  error_injection_rate ............................ 0
  error_injection_type ............................ transient_error
  eval_interval ................................... 1000
  eval_iters ...................................... 100
  evidence_data_path .............................. None
  exit_duration_in_mins ........................... None
  exit_interval ................................... None
  exit_on_missing_checkpoint ...................... False
  exit_signal_handler ............................. False
  exp_avg_dtype ................................... torch.float32
  exp_avg_sq_dtype ................................ torch.float32
  expert_model_parallel_size ...................... 1
  expert_tensor_parallel_size ..................... 1
  ffn_hidden_size ................................. 11008
  finetune ........................................ False
  flash_decode .................................... False
  fp16 ............................................ False
  fp16_lm_cross_entropy ........................... False
  fp32_residual_connection ........................ False
  fp8 ............................................. None
  fp8_amax_compute_algo ........................... most_recent
  fp8_amax_history_len ............................ 1
  fp8_interval .................................... 1
  fp8_margin ...................................... 0
  fp8_param_gather ................................ False
  fp8_wgrad ....................................... True
  global_batch_size ............................... 1024
  gradient_accumulation_fusion .................... True
  group_query_attention ........................... True
  head_lr_mult .................................... 1.0
  hidden_dropout .................................. 0.1
  hidden_size ..................................... 4096
  hierarchical_context_parallel_sizes ............. None
  hybrid_attention_ratio .......................... 0.0
  hybrid_mlp_ratio ................................ 0.0
  hybrid_override_pattern ......................... None
  hysteresis ...................................... 2
  ict_head_size ................................... None
  ict_load ........................................ None
  img_h ........................................... 224
  img_w ........................................... 224
  indexer_batch_size .............................. 128
  indexer_log_interval ............................ 1000
  inference_batch_times_seqlen_threshold .......... -1
  inference_max_seq_length ........................ 2560
  init_method_std ................................. 0.02
  init_method_xavier_uniform ...................... False
  initial_loss_scale .............................. 4294967296
  iter_per_epoch .................................. 1250
  iteration ....................................... 1
  kv_channels ..................................... 128
  kv_lora_rank .................................... 32
  lazy_mpu_init ................................... None
  load ............................................ /workdir/models/llama2-7b
  local_rank ...................................... 0
  log_interval .................................... 100
  log_loss_scale_to_tensorboard ................... True
  log_memory_to_tensorboard ....................... False
  log_num_zeros_in_grad ........................... False
  log_params_norm ................................. False
  log_progress .................................... False
  log_straggler ................................... False
  log_throughput .................................. False
  log_timers_to_tensorboard ....................... False
  log_validation_ppl_to_tensorboard ............... False
  log_world_size_to_tensorboard ................... False
  logging_level ................................... None
  loss_scale ...................................... None
  loss_scale_window ............................... 1000
  lr .............................................. None
  lr_decay_iters .................................. None
  lr_decay_samples ................................ None
  lr_decay_style .................................. linear
  lr_warmup_fraction .............................. None
  lr_warmup_init .................................. 0.0
  lr_warmup_iters ................................. 0
  lr_warmup_samples ............................... 0
  lr_wsd_decay_iters .............................. None
  lr_wsd_decay_samples ............................ None
  lr_wsd_decay_style .............................. exponential
  main_grads_dtype ................................ torch.float32
  main_params_dtype ............................... torch.float32
  make_vocab_size_divisible_by .................... 128
  manual_gc ....................................... False
  manual_gc_eval .................................. True
  manual_gc_interval .............................. 0
  mask_factor ..................................... 1.0
  mask_prob ....................................... 0.15
  mask_type ....................................... random
  masked_softmax_fusion ........................... False
  max_position_embeddings ......................... 4096
  max_tokens_to_oom ............................... 12000
  memory_snapshot_path ............................ snapshot.pickle
  merge_file ...................................... None
  micro_batch_size ................................ 1
  microbatch_group_size_per_vp_stage .............. None
  min_loss_scale .................................. 1.0
  min_lr .......................................... 0.0
  mmap_bin_files .................................. True
  mock_data ....................................... True
  moe_aux_loss_coeff .............................. 0.0
  moe_expert_capacity_factor ...................... None
  moe_extended_tp ................................. False
  moe_ffn_hidden_size ............................. 11008
  moe_grouped_gemm ................................ False
  moe_input_jitter_eps ............................ None
  moe_layer_freq .................................. 1
  moe_layer_recompute ............................. False
  moe_pad_expert_input_to_capacity ................ False
  moe_per_layer_logging ........................... False
  moe_router_load_balancing_type .................. aux_loss
  moe_router_pre_softmax .......................... False
  moe_router_topk ................................. 2
  moe_router_topk_limited_devices ................. None
  moe_router_topk_scaling_factor .................. None
  moe_shared_expert_intermediate_size ............. None
  moe_shared_expert_overlap ....................... False
  moe_token_dispatcher_type ....................... allgather
  moe_token_drop_policy ........................... probs
  moe_use_legacy_grouped_gemm ..................... False
  moe_use_upcycling ............................... False
  moe_z_loss_coeff ................................ None
  multi_latent_attention .......................... False
  nccl_communicator_config_path ................... None
  no_load_optim ................................... True
  no_load_rng ..................................... True
  no_persist_layer_norm ........................... False
  no_save_optim ................................... True
  no_save_rng ..................................... True
  non_persistent_ckpt_type ........................ None
  non_persistent_global_ckpt_dir .................. None
  non_persistent_local_ckpt_algo .................. fully_parallel
  non_persistent_local_ckpt_dir ................... None
  non_persistent_save_interval .................... None
  norm_epsilon .................................... 1e-05
  normalization ................................... RMSNorm
  num_attention_heads ............................. 32
  num_channels .................................... 3
  num_classes ..................................... 1000
  num_dataset_builder_threads ..................... 1
  num_distributed_optimizer_instances ............. 1
  num_experts ..................................... None
  num_layers ...................................... 32
  num_layers_per_virtual_pipeline_stage ........... None
  num_query_groups ................................ 32
  num_workers ..................................... 2
  one_logger_async ................................ False
  one_logger_project .............................. megatron-lm
  one_logger_run_name ............................. None
  onnx_safe ....................................... None
  openai_gelu ..................................... False
  optimizer ....................................... adam
  output_bert_embeddings .......................... False
  overlap_grad_reduce ............................. False
  overlap_p2p_comm ................................ False
  overlap_p2p_comm_warmup_flush ................... False
  overlap_param_gather ............................ False
  overlap_param_gather_with_optimizer_step ........ False
  override_opt_param_scheduler .................... False
  padded_vocab_size ............................... 32000
  params_dtype .................................... torch.float32
  patch_dim ....................................... 16
  per_split_data_args_path ........................ None
  perform_initialization .......................... False
  pipeline_model_parallel_size .................... 1
  pipeline_model_parallel_split_rank .............. None
  position_embedding_type ......................... rope
  pretrained_checkpoint ........................... None
  profile ......................................... False
  profile_ranks ................................... [0]
  profile_step_end ................................ 12
  profile_step_start .............................. 10
  q_lora_rank ..................................... None
  qk_head_dim ..................................... 128
  qk_layernorm .................................... False
  qk_pos_emb_head_dim ............................. 64
  query_in_block_prob ............................. 0.1
  rampup_batch_size ............................... None
  rank ............................................ 0
  recompute_granularity ........................... None
  recompute_method ................................ None
  recompute_num_layers ............................ None
  record_memory_history ........................... False
  renormalize_blend_weights ....................... False
  rerun_mode ...................................... disabled
  reset_attention_mask ............................ False
  reset_position_ids .............................. False
  retriever_report_topk_accuracies ................ []
  retriever_score_scaling ......................... False
  retriever_seq_length ............................ 256
  retro_add_retriever ............................. False
  retro_attention_gate ............................ 1
  retro_cyclic_train_iters ........................ None
  retro_encoder_attention_dropout ................. 0.1
  retro_encoder_hidden_dropout .................... 0.1
  retro_encoder_layers ............................ 2
  retro_num_neighbors ............................. 2
  retro_num_retrieved_chunks ...................... 2
  retro_project_dir ............................... None
  retro_verify_neighbor_count ..................... True
  rotary_base ..................................... 10000
  rotary_interleaved .............................. False
  rotary_percent .................................. 1.0
  rotary_scaling_factor ........................... 1.0
  rotary_seq_len_interpolation_factor ............. None
  s3_cache_path ................................... None
  sample_rate ..................................... 1.0
  save ............................................ None
  save_interval ................................... None
  scatter_gather_tensors_in_pipeline .............. True
  seed ............................................ 1234
  seq_length ...................................... 4096
  sequence_parallel ............................... False
  sgd_momentum .................................... 0.9
  short_seq_prob .................................. 0.1
  skip_train ...................................... False
  skipped_train_samples ........................... 0
  spec ............................................ None
  split ........................................... None
  squared_relu .................................... False
  standalone_embedding_stage ...................... False
  start_weight_decay .............................. 0.01
  straggler_ctrlr_port ............................ 65535
  straggler_minmax_count .......................... 1
  swiglu .......................................... True
  swin_backbone_type .............................. tiny
  tensor_model_parallel_size ...................... 1
  tensorboard_dir ................................. None
  tensorboard_log_interval ........................ 1
  tensorboard_queue_size .......................... 1000
  test_data_path .................................. None
  test_mode ....................................... False
  tiktoken_num_special_tokens ..................... 1000
  tiktoken_pattern ................................ None
  tiktoken_special_tokens ......................... None
  timing_log_level ................................ 0
  timing_log_option ............................... minmax
  titles_data_path ................................ None
  tokenizer_model ................................. /workdir/models/llama2-7b
  tokenizer_type .................................. Llama2Tokenizer
  tp_comm_bootstrap_backend ....................... nccl
  tp_comm_bulk_dgrad .............................. True
  tp_comm_bulk_wgrad .............................. True
  tp_comm_overlap ................................. False
  tp_comm_overlap_ag .............................. True
  tp_comm_overlap_cfg ............................. None
  tp_comm_overlap_rs .............................. True
  tp_comm_overlap_rs_dgrad ........................ False
  tp_comm_split_ag ................................ True
  tp_comm_split_rs ................................ True
  train_data_path ................................. None
  train_iters ..................................... None
  train_samples ................................... None
  train_sync_interval ............................. None
  transformer_impl ................................ transformer_engine
  transformer_pipeline_model_parallel_size ........ 1
  untie_embeddings_and_output_weights ............. True
  use_checkpoint_args ............................. False
  use_checkpoint_opt_param_scheduler .............. False
  use_cpu_initialization .......................... True
  use_dist_ckpt ................................... True
  use_dist_ckpt_deprecated ........................ False
  use_distributed_optimizer ....................... False
  use_flash_attn .................................. False
  use_legacy_models ............................... False
  use_mp_args_from_checkpoint_args ................ False
  use_one_sent_docs ............................... False
  use_precision_aware_optimizer ................... False
  use_pytorch_profiler ............................ False
  use_ring_exchange_p2p ........................... False
  use_rope_scaling ................................ False
  use_rotary_position_embeddings .................. False
  use_tokenizer_model_from_checkpoint_args ........ True
  use_torch_fsdp2 ................................. False
  use_tp_pp_dp_mapping ............................ False
  v_head_dim ...................................... 128
  valid_data_path ................................. None
  variable_seq_lengths ............................ False
  virtual_pipeline_model_parallel_size ............ None
  vision_backbone_type ............................ vit
  vision_pretraining .............................. False
  vision_pretraining_type ......................... classify
  vocab_extra_ids ................................. 0
  vocab_file ...................................... None
  vocab_size ...................................... 32000
  wandb_exp_name .................................. 
  wandb_project ................................... 
  wandb_save_dir .................................. 
  weight_decay .................................... 0.01
  weight_decay_incr_style ......................... constant
  wgrad_deferral_limit ............................ 0
  world_size ...................................... 1
  yaml_cfg ........................................ None
-------------------- end of arguments ---------------------
Loading checkpoint shards: 100%|█████████████████████████████████████| 2/2 [00:02<00:00,  1.05s/it]
building GPT model ...
set layer states: 100%|████████████████████████████████████████████| 32/32 [00:03<00:00,  8.98it/s]
sending embeddings
Overwriting default ffn_hidden_size value None with value from checkpoint 11008.
Overwriting default kv_channels value None with value from checkpoint 128.
Overwriting default group_query_attention value False with value from checkpoint True.
Overwriting default num_query_groups value 1 with value from checkpoint 32.
Overwriting default normalization value LayerNorm with value from checkpoint RMSNorm.
Overwriting default swiglu value False with value from checkpoint True.
Overwriting default global_batch_size value None with value from checkpoint 1024.
Overwriting default dataloader_type value None with value from checkpoint single.
Overwriting default use_legacy_models value False with value from checkpoint True.
Overwriting default load value None with value from checkpoint /workdir/models/llama2-7b.
Overwriting default overlap_p2p_comm value True with value from checkpoint False.
Overwriting default align_param_gather value True with value from checkpoint False.
Overwriting default mock_data value False with value from checkpoint True.
Overwriting default vocab_size value None with value from checkpoint 32000.
Overwriting default expert_tensor_parallel_size value None with value from checkpoint 1.
Overwriting default moe_ffn_hidden_size value None with value from checkpoint 11008.
Overwriting default transformer_impl value transformer_engine with value from checkpoint local.
Overwriting default main_grads_dtype value fp32 with value from checkpoint torch.float32.
Overwriting default main_params_dtype value fp32 with value from checkpoint torch.float32.
Overwriting default exp_avg_dtype value fp32 with value from checkpoint torch.float32.
Overwriting default exp_avg_sq_dtype value fp32 with value from checkpoint torch.float32.
Checkpoint had argument iteration but new arguments does not have this.
Checkpoint had argument padded_vocab_size but new arguments does not have this.
Checkpoint had argument use_dist_ckpt but new arguments does not have this.
Checkpoint had argument transformer_pipeline_model_parallel_size but new arguments does not have this.
Checkpoint had argument data_parallel_size but new arguments does not have this.
Checkpoint had argument consumed_train_samples but new arguments does not have this.
Checkpoint had argument skipped_train_samples but new arguments does not have this.
Checkpoint had argument consumed_valid_samples but new arguments does not have this.
Checkpoint had argument variable_seq_lengths but new arguments does not have this.
Checkpoint had argument disable_bias_linear but new arguments does not have this.
Checkpoint had argument model_type but new arguments does not have this.
Checkpoint had argument model_size but new arguments does not have this.
using world size: 1, data-parallel size: 1, context-parallel size: 1, hierarchical context-parallel sizes: Nonetensor-model-parallel size: 1, encoder-tensor-model-parallel size: 0, pipeline-model-parallel size: 1, encoder-pipeline-model-parallel size: 0
Process Process-1:
Traceback (most recent call last):
  File "/usr/lib/python3.12/multiprocessing/process.py", line 314, in _bootstrap
    self.run()
  File "/usr/lib/python3.12/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/workdir/Megatron-LM/tools/checkpoint/saver_megatron.py", line 168, in save_checkpoint
    validate_args(margs)
  File "/workdir/Megatron-LM/megatron/training/arguments.py", line 405, in validate_args
    args.main_grads_dtype = dtype_map[args.main_grads_dtype]
                            ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^
KeyError: torch.float32
sending transformer layer 0
sending transformer layer 1
sending transformer layer 2
sending transformer layer 3
sending transformer layer 4
sending transformer layer 5
sending transformer layer 6
sending transformer layer 7
sending transformer layer 8
sending transformer layer 9
sending transformer layer 10
sending transformer layer 11
sending transformer layer 12
sending transformer layer 13
sending transformer layer 14
sending transformer layer 15
sending transformer layer 16
sending transformer layer 17
sending transformer layer 18
sending transformer layer 19
sending transformer layer 20
sending transformer layer 21
sending transformer layer 22
sending transformer layer 23
sending transformer layer 24
sending transformer layer 25
sending transformer layer 26
sending transformer layer 27
sending transformer layer 28
sending transformer layer 29
sending transformer layer 30
sending transformer layer 31
sending final norm
sending output layer
Waiting for saver to complete...

Environment (please complete the following information):

• Docker images: nvcr.io/nvidia/pytorch 24.12-py3 https://catalog.ngc.nvidia.com/orgs/nvidia/containers/pytorch
• Megatron-LM commit ID commit 076972e
• PyTorch version 2.6.0a0+df5bbc09d1.nv24.12
• CUDA version cuda_12.6.r12.6/compiler.35059454_0
• transformers 4.47.1

Proposed fix
If you have a proposal for how to fix the issue state it here or link to a PR.

Additional context
Add any other context about the problem here.
When i try to install the transformers=4.31/4.32, there will be an error about compiling the tokenizer whl. So i choose to "pip install transformers" instead. I follow the guidence at https://github.com/NVIDIA/Megatron-LM/blob/main/docs/llama_mistral.md

[houshuai-cs]

i meet this question too!

[Sun2018421]

i meet this question too!

I am going to try Nemo for finetuning the model :)