See axolotl config

axolotl version: 0.4.1

base_model: IntervitensInc_gemma-2-27b-chatml
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

#trust_remote_code: true

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  #- path: anthracite-org/stheno-filtered-v1.1
  - path: stheno_data.json
    type: sharegpt
    conversation: chatml
      #- path: anthracite-org/kalo-opus-instruct-22k-no-refusal
  - path: kalo_opus_22k.jsonl
    type: sharegpt
    conversation: chatml
  #- path: anthracite-org/nopm_claude_writing_fixed
  - path: nopm_claude_dataset.jsonl
    type: sharegpt
    conversation: chatml
      #- path: Epiculous/Synthstruct-Gens-v1.1-Filtered-n-Cleaned
  - path: Epic_Synthstruct.json
    type: sharegpt
    conversation: chatml
      #- path: Epiculous/SynthRP-Gens-v1.1-Filtered-n-Cleaned
  - path: SynthRP-Gens_processed.json
    type: sharegpt
    conversation: chatml
chat_template: chatml
shuffle_merged_datasets: true
default_system_message: "You are an assistant that responds to the user."
dataset_prepared_path: magnum-v3-27b-data
val_set_size: 0.0
output_dir: ./magnum-v3-27b-r1

sequence_len: 8192
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len:

adapter:
lora_model_dir:
lora_r:
lora_alpha:
lora_dropout:
lora_target_linear:
lora_fan_in_fan_out:

wandb_project: magnum-v3-27b-r1
wandb_entity:
wandb_watch:
wandb_name: attempt-01
wandb_log_model:

gradient_accumulation_steps: 8
micro_batch_size: 1
num_epochs: 2
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 0.000004

plugins:                                                                                                                                                              
  - axolotl.integrations.liger.LigerPlugin                                                                                                                            
liger_cross_entropy: true
    #liger_rope: true
    #liger_rms_norm: true
    #liger_swiglu: true
    #liger_fused_linear_cross_entropy: true

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


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

warmup_steps: 40
evals_per_epoch:
eval_table_size:
eval_max_new_tokens:
saves_per_epoch: 2
debug:
deepspeed: /dev/shm/axolotl/deepspeed_configs/zero3_bf16.json
weight_decay: 0.03
fsdp:
#  - full_shard
#  - auto_wrap
fsdp_config:
#  fsdp_limit_all_gathers: true
#  fsdp_sync_module_states: true
#  fsdp_offload_params: true
#  fsdp_use_orig_params: false
#  fsdp_cpu_ram_efficient_loading: false
#  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
#  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
#  fsdp_state_dict_type: FULL_STATE_DICT
special_tokens:
  pad_token: "<pad>"

magnum-v3-27b-r1

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: 4e-06
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• total_eval_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 40
• num_epochs: 2

Training results

Framework versions

• Transformers 4.43.2
• Pytorch 2.4.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1