--- library_name: transformers base_model: Columbidae/apertus-12b-chatml-untrained tags: - axolotl - generated_from_trainer datasets: - allura-org/the-anarchist-library - ToastyPigeon/steve-and-marvin - ToastyPigeon/kimi-stories-completion - ToastyPigeon/some-erotica - Alfitaria/rosier-inf - ToastyPigeon/SpringDragon - ToastyPigeon/erotic-books-clone model-index: - name: apertus-12b-cpt-attempt results: [] --- [Built with Axolotl](https://github.com/axolotl-ai-cloud/axolotl)
See axolotl config axolotl version: `0.12.2` ```yaml # === Model Configuration === base_model: Columbidae/apertus-12b-chatml-untrained trust_remote_code: true load_in_8bit: false load_in_4bit: false # === HF Configuration === hub_model_id: allura-forge/apertus-12b-cpt-attempt hub_strategy: "every_save" output_dir: ckpts # === Wandb Tracking === wandb_project: ApertusV2 # wandb_entity: [WANDB_ENTITY] wandb_name: 12b-cpt-part1 # === Training Setup === num_epochs: 1 micro_batch_size: 1 gradient_accumulation_steps: 16 sequence_len: 8192 #sequence_parallel_degree: 2 #heads_k_stride: 1 sample_packing: true #pad_to_sequence_len: true #temperature: 0.7 #max_steps: 10 # === Evaluation === #val_set_size: 0.025 #evals_per_epoch: 10 #eval_steps: 20 #max_steps: 60 #eval_table_size: #eval_max_new_tokens: 128 #eval_sample_packing: true eval_strategy: "no" # === LoRA Configuration === adapter: lora_model_dir: lora_r: 128 lora_alpha: 16 lora_dropout: 0.05 lora_target_linear: true lora_target_modules: # - up_proj # - down_proj # - gate_proj # - q_proj # - v_proj # - k_proj # - o_proj # - input_layernorm # - post_attention_layernorm # - embed_tokens # - lm_head lora_fan_in_fan_out: peft_use_rslora: true lora_modules_to_save: # - embed_tokens # - lm_head #fix_untrained_tokens: true #lora_mlp_kernel: true #lora_qkv_kernel: true #lora_o_kernel: true #unfrozen_parameters: # - model.layers.[0-9+].mlp.up_proj # - model.layers.[0-9+].mlp.down_proj # - model.layers.[0-9+].feedforward_layernorm # - embed_tokens # - lm_head # === Hyperparameter Configuration === #optimizer: apollo_adamw_layerwise #warmup_steps: 0 warmup_ratio: 0.025 #optimizer: adamw_torch_fused optimizer: paged_adamw_8bit #optim_args: # enable_stochastic_rounding: true # enable_cautious: true # enable_8bit: true # Apollo-mini configuration: #optim_args: "proj=random,rank=128,scale=128.0,scale_type=tensor,update_proj_gap=100" # Regular Apollo configuration: # optim_args: #optim_target_modules: all_linear learning_rate: 1e-5 lr_scheduler: cosine #cosine_min_lr_ratio: 0.2 #lr_scheduler: cosine_with_min_lr #lr_scheduler_kwargs: # cosine_min_lr: 1e-6 weight_decay: 0.01 max_grad_norm: 1.0 #warmup_steps: 0 #warmup_ratio: 0.025 # === Data Configuration === # #chat_template: jinja chat_template: chatml special_tokens: # eos_token: "<|im_end|>" # eos_token: "" #tokenizer_use_mistral_common: true shuffle_merged_datasets: true datasets: - path: allura-org/the-anarchist-library type: completion split: train[:20%] # - path: grimulkan/LimaRP-augmented # type: chat_template # field_messages: conversations # message_property_mappings: # role: from # content: value # - path: allenai/tulu-3-sft-personas-instruction-following # type: chat_template # split: train[:10%] # - path: ToastyPigeon/mixed-medical-reasoning-formatted # type: chat_template # data_files: mixed-medical-thinking.json # split: train[:10%] - path: ToastyPigeon/steve-and-marvin type: completion data_files: marvin.json - path: ToastyPigeon/kimi-stories-completion type: completion - path: ToastyPigeon/some-erotica # type: customcompletion-regex type: completion split: train[50%:65%] # data_files: new-story-dataset-v2.json # - path: allura-org/fujin-instruct-v2 # type: customchatml-regex # type: chat_template # field_messages: conversations # message_property_mappings: # role: from # content: value # - path: ToastyPigeon/some-rp-extended # type: customchatml-regex # type: chat_template # field_messages: conversations # message_property_mappings: # role: from # content: value # roles_to_train: ["user","assistant"] - path: Alfitaria/rosier-inf type: completion split: train[:20%] # - path: allura-forge/koto-instruct-sft-nothink # type: customchatml-regex # type: chat_template # split: train[:50%] # field_messages: conversations # message_property_mappings: # role: from # content: value - path: ToastyPigeon/SpringDragon # type: customcompletion-regex type: completion split: train - path: ToastyPigeon/erotic-books-clone # type: customcompletion-regex type: completion split: train[:20%] # split: train[35%:45%] # - path: ToastyPigeon/tulu-mini # type: chat_template dataset_prepared_path: last_run_prepared # === Plugins === plugins: - axolotl.integrations.liger.LigerPlugin - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin # === Hardware Optimization === gradient_checkpointing: true liger_rope: true liger_rms_norm: true liger_layer_norm: true liger_glu_activation: true #liger_fused_linear_cross_entropy: true cut_cross_entropy: true #deepspeed: ../axolotl/deepspeed_configs/zero2.json # === FSDP Config === #fsdp: # - full_shard # - auto_wrap #fsdp_config: # fsdp_limit_all_gathers: true # fsdp_sync_module_states: true # fsdp_offload_params: true # fsdp_activation_checkpointing: true # fsdp_use_orig_params: true # fsdp_cpu_ram_efficient_loading: true # fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP # fsdp_transformer_layer_cls_to_wrap: ApertusDecoderLayer # fsdp_state_dict_type: FULL_STATE_DICT # fsdp_sharding_strategy: FULL_SHARD #fsdp_stage: 2 #fsdp_final_state_dict_type: FULL_STATE_DICT # === Checkpointing === #save_steps: 2 saves_per_epoch: 10 save_total_limit: 1 # === Advanced Settings === bf16: auto flash_attention: true train_on_inputs: false group_by_length: false save_safetensors: true logging_steps: 1 seed: 420 gc_steps: 10 ```

# apertus-12b-cpt-attempt This model is a fine-tuned version of [Columbidae/apertus-12b-chatml-untrained](https://huggingface.co/Columbidae/apertus-12b-chatml-untrained) on the allura-org/the-anarchist-library, the ToastyPigeon/steve-and-marvin, the ToastyPigeon/kimi-stories-completion, the ToastyPigeon/some-erotica, the Alfitaria/rosier-inf, the ToastyPigeon/SpringDragon and the ToastyPigeon/erotic-books-clone datasets. ## 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: 1e-05 - train_batch_size: 1 - eval_batch_size: 1 - seed: 420 - distributed_type: multi-GPU - num_devices: 2 - gradient_accumulation_steps: 16 - total_train_batch_size: 32 - total_eval_batch_size: 2 - optimizer: Use 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: 7 - training_steps: 290 ### Training results ### Framework versions - Transformers 4.57.1 - Pytorch 2.8.0+cu129 - Datasets 4.0.0 - Tokenizers 0.22.1