Modelpublic

sleepdeprived3/Baptist-Christian-Bible-Expert-v2.0-22B-LoRA

sourceHugging Faceotherupdated 1y agoView on Hugging Face

0likes4downloads

Model Card

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

axolotl version: 0.8.1

yaml

base_model: mistralai/Mistral-Small-Instruct-2409
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

model_config:
  trust_remote_code: true
  tokenizer:
    pad_token: "</s>"
    padding_side: "right"
    add_bos_token: true
    add_eos_token: false

datasets:
  - path: data/data.jsonl
    type: chat_template
    chat_template_strategy: tokenizer
    field_messages: conversations
    message_property_mappings:
      role: role
      content: content
    roles:
      user: ["user"]
      assistant: ["assistant"]
      system: ["system"]

load_in_4bit: true
adapter: qlora
lora_r: 64
lora_alpha: 32
lora_dropout: 0.1
lora_target_modules:
  - q_proj
  - k_proj
  - v_proj
  - o_proj
  - gate_proj
  - up_proj
  - down_proj

bf16: true
flash_attention: true
gradient_checkpointing: true
deepspeed: deepspeed_configs/zero2.json

gradient_accumulation_steps: 4
micro_batch_size: 8
num_epochs: 2
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 3e-6
warmup_ratio: 0.02

max_seq_length: 8192
pad_to_sequence_len: true
sample_packing: true
max_grad_norm: 1.0

output_dir: ./output
save_steps: 100
logging_steps: 10
save_safetensors: true

special_tokens:
  pad_token: "</s>"

</details><br>

output

This model is a fine-tuned version of mistralai/Mistral-Small-Instruct-2409 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: 3e-06
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—distributed_type: multi-GPU
—gradientaccumulationsteps: 4
—totaltrainbatch_size: 32
—optimizer: Use pagedadamw8bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: cosine
—lrschedulerwarmup_steps: 8
—num_epochs: 2.0

Training results

Framework versions

—PEFT 0.15.1
—Transformers 4.51.0
—Pytorch 2.6.0+cu124
—Datasets 3.5.0
—Tokenizers 0.21.1