Edens-Gate/Baldur-Pleias-r1

See axolotl config

axolotl version: 0.5.2

base_model: ./pleias_erebus
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

load_in_8bit: false
load_in_4bit: false
strict: false


  #datasets:
  #  - path: Mielikki/Erebus-87k
  #    type: completion
  #    field: body
  #  - path: allura-org/r_shortstories_24k
  #    type: completion
  #    field: text
datasets:
  - path: Gryphe/Sonnet3.5-SlimOrcaDedupCleaned-20k
  - path: anthracite-org/kalo-opus-instruct-22k-no-refusal
  - path: anthracite-org/kalo_misc_part2
  - path: anthracite-org/kalo_opus_misc_240827
  - path: Nitral-AI/Olympiad_Math-ShareGPT
  - path: Nitral-AI/Cybersecurity-ShareGPT
  - path: Nitral-AI/Medical_Instruct-ShareGPT
  - path: NewEden/Claude-Instruct-2.7K
  - path: NewEden/Claude-Instruct-5K
dataset_config:
  type: chat_template
  field_messages: conversations
  message_field_role: from
  message_field_content: value
  roles_to_train: ["gpt"]
  train_on_eos: "turn"
#chat_template: jinja
  #chat_template_jinja: "{{ bos_token }}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<|START_OF_TURN_TOKEN|><|USER_TOKEN|>' + content + '<|END_OF_TURN_TOKEN|>' }}{% elif message['role'] == 'assistant' %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>'  + content + '<|END_OF_TURN_TOKEN|>' }}{% elif message['role'] == 'system' %}{{ '<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>'  + content + '<|END_OF_TURN_TOKEN|>' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>' }}{% endif %}"
chat_template: chatml

output_dir: ./pleias_outputs

sequence_len: 8192
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true

wandb_project: Pleias-Baldur
wandb_entity:
wandb_watch:
wandb_name: run-2
wandb_log_model:

plugins:
  - axolotl.integrations.liger.LigerPlugin
liger_rope: true
liger_layer_norm: true
liger_swiglu: true
liger_fused_linear_cross_entropy: true

gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 3
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 8e-5


train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false

  #gradient_checkpointing: unsloth
gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
early_stopping_patience:
resume_from_checkpoint:
#auto_resume_from_checkpoints: true
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 25
evals_per_epoch: 2
eval_table_size:
eval_max_new_tokens: 128
saves_per_epoch: 1
weight_decay: 0.02
debug:
  #deepspeed: /workspace/axolotl/deepspeed_configs/zero3_bf16_cpuoffload_params.json
deepspeed: /workspace/axolotl/deepspeed_configs/zero3_bf16.json
  #deepspeed: /workspace/axolotl/deepspeed_configs/zero2.json
fsdp:
fsdp_config:
special_tokens:
  bos_token: '<|begin_of_text|>'
  eos_token: '<|end_of_text|>'
  pad_token: '[PAD]'

pleias_outputs

This model was trained from scratch on the None dataset.

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 8e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• total_eval_batch_size: 8
• optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 25
• num_epochs: 3

Training results

Framework versions

• Transformers 4.46.3
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3