---
library_name: peft
tags:
- generated_from_trainer
datasets:
- GuilhermeNaturaUmana/Reasoning-deepseek
base_model: nicoboss/Hermes-3-Llama-3.1-405B-Uncensored
model-index:
- name: workspace/data/Hermes-3-Llama-3.1-405B-Uncensored-Reasoner
  results: []
---

Yes, its fine we have finally AGI Lora at home

AGI internally achieved confirmed lol(and now public)

now besides joke, this is an model made to reason, i didnt made the benchmark yet, but its gonna or be comparable to deepseek R1 or surpasses it!
(the merged model is on the works and its gonna be released today!)

 --------------
 
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

[<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
<details><summary>See axolotl config</summary>

axolotl version: `0.6.0`
```yaml
base_model: /root/Hermes-3-Llama-3.1-405B-Uncensored
tokenizer_type: AutoTokenizer

load_in_4bit: true
strict: false

datasets:
  - path: GuilhermeNaturaUmana/Reasoning-deepseek
    type: chat_template
    chat_template: llama3
    field_messages: messages
    message_field_role: role
    message_field_content: content
    roles:
      system:
        - system
      user:
        - user
      assistant:
        - assistant
dataset_prepared_path: last_run_prepared
val_set_size: 0.0
output_dir: /workspace/data/Hermes-3-Llama-3.1-405B-Uncensored-Reasoner
save_safetensors: true

adapter: qlora

sequence_len: 2048
sample_packing: true
pad_to_sequence_len: true

lora_r: 16
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
lora_target_linear: true

gradient_accumulation_steps: 4
micro_batch_size: 1
num_epochs: 1
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 0.00001

train_on_inputs: false
group_by_length: false
bf16: true
tf32: true

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: true
logging_steps: 1
flash_attention: true

warmup_steps: 10
evals_per_epoch: 6
saves_per_epoch: 6
save_total_limit: 20
weight_decay: 0.0
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: true
  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
  fsdp_state_dict_type: FULL_STATE_DICT
  fsdp_sharding_strategy: FULL_SHARD
special_tokens:
  pad_token: <|finetune_right_pad_id|>

```

</details><br>

# workspace/data/Hermes-3-Llama-3.1-405B-Uncensored-Reasoner

This model was trained from scratch on the GuilhermeNaturaUmana/Reasoning-deepseek 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: 1e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 3
- gradient_accumulation_steps: 4
- total_train_batch_size: 12
- total_eval_batch_size: 3
- optimizer: Use OptimizerNames.ADAMW_TORCH 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: 10
- num_epochs: 1.0

### Training results



### Framework versions

- PEFT 0.14.0
- Transformers 4.48.1
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.0