---
language: pt
library_name: peft
datasets: Weni/zeroshot-3.0.3
pipeline_tag: zero-shot-classification
training_arguments:
  output_dir: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/
  overwrite_output_dir: false
  do_train: false
  do_eval: true
  do_predict: false
  evaluation_strategy: epoch
  prediction_loss_only: false
  per_device_train_batch_size: 2
  per_device_eval_batch_size: 8
  gradient_accumulation_steps: 2
  eval_accumulation_steps: 1
  eval_delay: 0
  learning_rate: 0.0004
  weight_decay: 0.01
  adam_beta1: 0.9
  adam_beta2: 0.999
  adam_epsilon: 1.0e-08
  max_grad_norm: 0.3
  num_train_epochs: 10
  max_steps: -1
  lr_scheduler_type: cosine
  warmup_ratio: 0.1
  warmup_steps: 0
  log_level: passive
  log_level_replica: warning
  log_on_each_node: true
  logging_dir: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/runs/Dec01_21-53-07_fd10189bb234
  logging_strategy: steps
  logging_first_step: false
  logging_steps: 500
  logging_nan_inf_filter: true
  save_strategy: epoch
  save_steps: 500
  save_total_limit: 5
  save_safetensors: true
  save_on_each_node: false
  no_cuda: false
  use_mps_device: false
  seed: 42
  jit_mode_eval: false
  use_ipex: false
  bf16: false
  fp16: true
  fp16_opt_level: O1
  half_precision_backend: auto
  bf16_full_eval: false
  fp16_full_eval: false
  local_rank: 0
  tpu_metrics_debug: false
  debug: []
  dataloader_drop_last: false
  dataloader_num_workers: 0
  past_index: -1
  run_name: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/
  disable_tqdm: false
  remove_unused_columns: true
  load_best_model_at_end: true
  metric_for_best_model: eval_loss
  greater_is_better: false
  ignore_data_skip: false
  sharded_ddp: []
  fsdp: []
  fsdp_min_num_params: 0
  fsdp_config:
    fsdp_min_num_params: 0
    xla: false
    xla_fsdp_grad_ckpt: false
  label_smoothing_factor: 0.0
  optim: adamw_torch
  adafactor: false
  group_by_length: false
  length_column_name: length
  report_to:
  - tensorboard
  dataloader_pin_memory: true
  skip_memory_metrics: true
  use_legacy_prediction_loop: false
  push_to_hub: true
  hub_model_id: Weni/ZeroShot-2.2.1-Llama2-13b-Multilanguage-3.0.3
  hub_strategy: all_checkpoints
  hub_token: <HUB_TOKEN>
  hub_private_repo: false
  gradient_checkpointing: true
  include_inputs_for_metrics: false
  fp16_backend: auto
  push_to_hub_token: <PUSH_TO_HUB_TOKEN>
  mp_parameters: ''
  auto_find_batch_size: false
  full_determinism: false
  ray_scope: last
  ddp_timeout: 1800
  torch_compile: false
datatset:
  name: Weni/zeroshot-3.0.3
Training Procedure:
  Training Hyperparameters:
    output_dir: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/
    overwrite_output_dir: false
    do_train: false
    do_eval: true
    do_predict: false
    evaluation_strategy: epoch
    prediction_loss_only: false
    per_device_train_batch_size: 2
    per_device_eval_batch_size: 8
    gradient_accumulation_steps: 2
    eval_accumulation_steps: 1
    eval_delay: 0
    learning_rate: 0.0004
    weight_decay: 0.01
    adam_beta1: 0.9
    adam_beta2: 0.999
    adam_epsilon: 1.0e-08
    max_grad_norm: 0.3
    num_train_epochs: 10
    max_steps: -1
    lr_scheduler_type: cosine
    warmup_ratio: 0.1
    warmup_steps: 0
    log_level: passive
    log_level_replica: warning
    log_on_each_node: true
    logging_dir: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/runs/Dec01_21-53-07_fd10189bb234
    logging_strategy: steps
    logging_first_step: false
    logging_steps: 500
    logging_nan_inf_filter: true
    save_strategy: epoch
    save_steps: 500
    save_total_limit: 5
    save_safetensors: true
    save_on_each_node: false
    no_cuda: false
    use_mps_device: false
    seed: 42
    jit_mode_eval: false
    use_ipex: false
    bf16: false
    fp16: true
    fp16_opt_level: O1
    half_precision_backend: auto
    bf16_full_eval: false
    fp16_full_eval: false
    local_rank: 0
    tpu_metrics_debug: false
    debug: []
    dataloader_drop_last: false
    dataloader_num_workers: 0
    past_index: -1
    run_name: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/
    disable_tqdm: false
    remove_unused_columns: true
    load_best_model_at_end: true
    metric_for_best_model: eval_loss
    greater_is_better: false
    ignore_data_skip: false
    sharded_ddp: []
    fsdp: []
    fsdp_min_num_params: 0
    fsdp_config:
      fsdp_min_num_params: 0
      xla: false
      xla_fsdp_grad_ckpt: false
    label_smoothing_factor: 0.0
    optim: adamw_torch
    adafactor: false
    group_by_length: false
    length_column_name: length
    report_to:
    - tensorboard
    dataloader_pin_memory: true
    skip_memory_metrics: true
    use_legacy_prediction_loop: false
    push_to_hub: true
    hub_model_id: Weni/ZeroShot-2.2.1-Llama2-13b-Multilanguage-3.0.3
    hub_strategy: all_checkpoints
    hub_token: <HUB_TOKEN>
    hub_private_repo: false
    gradient_checkpointing: true
    include_inputs_for_metrics: false
    fp16_backend: auto
    push_to_hub_token: <PUSH_TO_HUB_TOKEN>
    mp_parameters: ''
    auto_find_batch_size: false
    full_determinism: false
    ray_scope: last
    ddp_timeout: 1800
    torch_compile: false
  Training data:
    name: Weni/zeroshot-3.0.3
  Training processing: 'dataset = dataset.shuffle(seed=55)

    dataset = dataset[''train''].train_test_split(test_size=0.1)'
training_regime: "### Training Hyperparameters- output_dir: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/\n\
  - overwrite_output_dir: False\n- do_train: False\n- do_eval: True\n- do_predict:\
  \ False\n- evaluation_strategy: epoch\n- prediction_loss_only: False\n- per_device_train_batch_size:\
  \ 2\n- per_device_eval_batch_size: 8\n- per_gpu_train_batch_size: None\n- per_gpu_eval_batch_size:\
  \ None\n- gradient_accumulation_steps: 2\n- eval_accumulation_steps: 1\n- eval_delay:\
  \ 0\n- learning_rate: 0.0004\n- weight_decay: 0.01\n- adam_beta1: 0.9\n- adam_beta2:\
  \ 0.999\n- adam_epsilon: 1e-08\n- max_grad_norm: 0.3\n- num_train_epochs: 10\n-\
  \ max_steps: -1\n- lr_scheduler_type: cosine\n- warmup_ratio: 0.1\n- warmup_steps:\
  \ 0\n- log_level: passive\n- log_level_replica: warning\n- log_on_each_node: True\n\
  - logging_dir: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/runs/Dec01_21-53-07_fd10189bb234\n\
  - logging_strategy: steps\n- logging_first_step: False\n- logging_steps: 500\n-\
  \ logging_nan_inf_filter: True\n- save_strategy: epoch\n- save_steps: 500\n- save_total_limit:\
  \ 5\n- save_safetensors: True\n- save_on_each_node: False\n- no_cuda: False\n- use_mps_device:\
  \ False\n- seed: 42\n- data_seed: None\n- jit_mode_eval: False\n- use_ipex: False\n\
  - bf16: False\n- fp16: True\n- fp16_opt_level: O1\n- half_precision_backend: auto\n\
  - bf16_full_eval: False\n- fp16_full_eval: False\n- tf32: None\n- local_rank: 0\n\
  - ddp_backend: None\n- tpu_num_cores: None\n- tpu_metrics_debug: False\n- debug:\
  \ []\n- dataloader_drop_last: False\n- eval_steps: None\n- dataloader_num_workers:\
  \ 0\n- past_index: -1\n- run_name: ./Zeroshot/01-12-23-NousResearch-Nous-Hermes-Llama2-13b_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_2/checkpoints/\n\
  - disable_tqdm: False\n- remove_unused_columns: True\n- label_names: None\n- load_best_model_at_end:\
  \ True\n- metric_for_best_model: eval_loss\n- greater_is_better: False\n- ignore_data_skip:\
  \ False\n- sharded_ddp: []\n- fsdp: []\n- fsdp_min_num_params: 0\n- fsdp_config:\
  \ {'fsdp_min_num_params': 0, 'xla': False, 'xla_fsdp_grad_ckpt': False}\n- fsdp_transformer_layer_cls_to_wrap:\
  \ None\n- deepspeed: None\n- label_smoothing_factor: 0.0\n- optim: adamw_torch\n\
  - optim_args: None\n- adafactor: False\n- group_by_length: False\n- length_column_name:\
  \ length\n- report_to: ['tensorboard']\n- ddp_find_unused_parameters: None\n- ddp_bucket_cap_mb:\
  \ None\n- ddp_broadcast_buffers: None\n- dataloader_pin_memory: True\n- skip_memory_metrics:\
  \ True\n- use_legacy_prediction_loop: False\n- push_to_hub: True\n- resume_from_checkpoint:\
  \ None\n- hub_model_id: Weni/ZeroShot-2.2.1-Llama2-13b-Multilanguage-3.0.3\n- hub_strategy:\
  \ all_checkpoints\n- hub_token: <HUB_TOKEN>\n- hub_private_repo: False\n- gradient_checkpointing:\
  \ True\n- include_inputs_for_metrics: False\n- fp16_backend: auto\n- push_to_hub_model_id:\
  \ None\n- push_to_hub_organization: None\n- push_to_hub_token: <PUSH_TO_HUB_TOKEN>\n\
  - mp_parameters: \n- auto_find_batch_size: False\n- full_determinism: False\n- torchdynamo:\
  \ None\n- ray_scope: last\n- ddp_timeout: 1800\n- torch_compile: False\n- torch_compile_backend:\
  \ None\n- torch_compile_mode: None\n- xpu_backend: None"
training_data:
  name: Weni/zeroshot-3.0.3
'preprocessing ': 'dataset = dataset.shuffle(seed=55)

  dataset = dataset[''train''].train_test_split(test_size=0.1)'
preprocessing: 'dataset = dataset.shuffle(seed=55)

  dataset = dataset[''train''].train_test_split(test_size=0.1)'
base_model: NousResearch/Nous-Hermes-Llama2-13b
---

## Training Hyperparameters
- evaluation_strategy: epoch
- prediction_loss_only: False
- per_device_train_batch_size: 2
- per_device_eval_batch_size: 8
- per_gpu_train_batch_size: None
- per_gpu_eval_batch_size: None
- gradient_accumulation_steps: 2
- eval_accumulation_steps: 1
- eval_delay: 0
- learning_rate: 0.0004
- weight_decay: 0.01
- adam_beta1: 0.9
- adam_beta2: 0.999
- adam_epsilon: 1e-08
- max_grad_norm: 0.3
- num_train_epochs: 10
- max_steps: -1
- lr_scheduler_type: cosine
- warmup_ratio: 0.1
- warmup_steps: 0
- log_level: passive
- log_level_replica: warning
- log_on_each_node: True
- logging_strategy: steps
- logging_first_step: False
- logging_steps: 500
- logging_nan_inf_filter: True
- save_strategy: epoch
- save_steps: 500
- save_total_limit: 5
- save_safetensors: True
- save_on_each_node: False
- no_cuda: False
- use_mps_device: False
- seed: 42
- data_seed: None
- jit_mode_eval: False
- use_ipex: False
- bf16: False
- fp16: True
- fp16_opt_level: O1
- half_precision_backend: auto
- bf16_full_eval: False
- fp16_full_eval: False
- tf32: None
- local_rank: 0
- ddp_backend: None
- tpu_num_cores: None
- tpu_metrics_debug: False
- debug: []
- dataloader_drop_last: False
- eval_steps: None
- dataloader_num_workers: 0
- past_index: -1
- remove_unused_columns: True
- label_names: None
- load_best_model_at_end: True
- metric_for_best_model: eval_loss
- greater_is_better: False
- ignore_data_skip: False
- sharded_ddp: []
- fsdp: []
- fsdp_min_num_params: 0
- fsdp_config: {'fsdp_min_num_params': 0, 'xla': False, 'xla_fsdp_grad_ckpt': False}
- fsdp_transformer_layer_cls_to_wrap: None
- deepspeed: None
- label_smoothing_factor: 0.0
- optim: adamw_torch
- optim_args: None
- adafactor: False
- group_by_length: False
- ddp_find_unused_parameters: None
- ddp_bucket_cap_mb: None
- ddp_broadcast_buffers: None
- dataloader_pin_memory: True
- skip_memory_metrics: True
- use_legacy_prediction_loop: False
- push_to_hub: True
- resume_from_checkpoint: None
- hub_strategy: all_checkpoints
- gradient_checkpointing: True
    
## Training procedure


The following `bitsandbytes` quantization config was used during training:
- load_in_8bit: False
- load_in_4bit: True
- llm_int8_threshold: 6.0
- llm_int8_skip_modules: None
- llm_int8_enable_fp32_cpu_offload: False
- llm_int8_has_fp16_weight: False
- bnb_4bit_quant_type: nf4
- bnb_4bit_use_double_quant: True
- bnb_4bit_compute_dtype: bfloat16

- 
### Framework versions


- PEFT 0.4.0