See axolotl config

axolotl version: 0.4.1

base_model: mistralai/Mistral-7B-Instruct-v0.3
# optionally might have model_type or tokenizer_type
model_type: MistralForCausalLM
tokenizer_type: LlamaTokenizer
# Automatically upload checkpoint and final model to HF
hub_model_id: AiAF/Finetuned-SCPWiki-032025-7B-Instruct

load_in_8bit: false
load_in_4bit: true
strict: false

datasets:
  - path: AiAF/Finetuning-SCPWiki-032025-7B-Instruct-plain_qa_list
    ds_type: json
    type: chat_template
    chat_template: chatml
    field_messages: conversations
    message_field_role: from
    message_field_content: value
    roles:
      user:
        - human
      assistant:
        - gpt
      system:
        - system
        
dataset_prepared_path: last_run_prepared
val_set_size: 0.1
output_dir: ./outputs/qlora-out/Finetuned-SCPWiki-032025-7B-Instruct-V1

adapter: qlora
lora_model_dir:

sequence_len: 1024
sample_packing: true
pad_to_sequence_len: true

lora_r: 16
lora_alpha: 32
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_proj

wandb_project: "LLM-Finetuning"
wandb_entity:
wandb_watch: "all"
wandb_name: "Finetuned-SCPWiki-032025-7B-Instruct-V1"
wandb_log_model: "false"

gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 4
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.000005

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

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

loss_watchdog_threshold: 5.0
loss_watchdog_patience: 3

warmup_steps: 10
evals_per_epoch: 5
eval_table_size:
eval_max_new_tokens: 128
saves_per_epoch: 5
debug:
deepspeed: deepspeed_configs/zero2.json
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:

Finetuned-SCPWiki-032025-7B-Instruct

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2898

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: 5e-06
• 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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss
1.3436	0.0301	1	1.3905
1.358	0.2105	7	1.3875
1.2051	0.4211	14	1.3651
1.3802	0.6316	21	1.3418
1.266	0.8421	28	1.3300
1.287	1.0526	35	1.3222
1.2226	1.2632	42	1.3141
1.1621	1.4737	49	1.3074
1.4156	1.6842	56	1.3021
1.3251	1.8947	63	1.2985
1.0981	2.1053	70	1.2956
1.3638	2.3158	77	1.2936
1.1033	2.5263	84	1.2923
1.2248	2.7368	91	1.2911
1.2687	2.9474	98	1.2903
1.186	3.1579	105	1.2900
1.325	3.3684	112	1.2896
1.2011	3.5789	119	1.2896
1.1852	3.7895	126	1.2898

Framework versions

• PEFT 0.13.2
• Transformers 4.45.2
• Pytorch 2.3.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.1