FourOhFour/Fatgirl_8B

See axolotl config

axolotl version: 0.4.1

base_model: jeiku/Magic_8B
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: anthracite-org/stheno-filtered-v1.1
    type: sharegpt
    conversation: chatml
  - path: Epiculous/SynthRP-Gens-v1.1-Filtered-n-Cleaned
    type: sharegpt
    conversation: chatml
  - path: ResplendentAI/bluemoon
    type: sharegpt
    conversation: chatml
  - path: openerotica/freedom-rp
    type: sharegpt
    conversation: chatml
  - path: MinervaAI/Aesir-Preview
    type: sharegpt
    conversation: chatml

chat_template: chatml

val_set_size: 0.01
output_dir: ./outputs/out

adapter:
lora_r:
lora_alpha:
lora_dropout:
lora_target_linear:

sequence_len: 8192
# sequence_len: 32768
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true

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

wandb_project: New8B
wandb_entity:
wandb_watch:
wandb_name: New8B
wandb_log_model:

gradient_accumulation_steps: 32
micro_batch_size: 1
num_epochs: 2
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.00001
weight_decay: 0.05

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

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

warmup_ratio: 0.1
evals_per_epoch: 4
eval_table_size:
eval_max_new_tokens: 128
saves_per_epoch: 2

debug:
deepspeed:
fsdp:
fsdp_config:

special_tokens:
  pad_token: <pad>

outputs/out

This model is a fine-tuned version of jeiku/Magic_8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3029

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: 2
• gradient_accumulation_steps: 32
• total_train_batch_size: 64
• total_eval_batch_size: 2
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 32
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
1.447	0.0062	1	1.4349
1.3437	0.2530	41	1.3502
1.3734	0.5060	82	1.3237
1.3543	0.7590	123	1.3128
1.319	1.0102	164	1.3064
1.2886	1.2636	205	1.3042
1.2387	1.5169	246	1.3031
1.3746	1.7702	287	1.3029

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1