You can deploy this model and start earning money today!

alexandra-e/
t-lite-stalin

Name: t-lite-stalin
Availability: InStock

peft

safetensors

llama

generated_from_trainer

AnatoliiPotapov/T-lite-instruct-0.1

adapter

4-bit

bitsandbytes

Views

No views yet

See axolotl config

axolotl version: 0.4.1

yaml

base_model: t-bank-ai/T-lite-instruct-0.1

load_in_8bit: false
load_in_4bit: true
strict: false

datasets:
  - path: test.jsonl
    type: completion
dataset_prepared_path: prepared_data_tlite
val_set_size: 0.1
output_dir: ./t-lite-stalin

adapter: qlora
lora_model_dir:

sequence_len: 272
sample_packing: true
eval_sample_packing: False
pad_to_sequence_len: true

lora_r: 32
lora_alpha: 16
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: 
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 48
num_epochs: 4
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0002

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: 1
eval_table_size:
eval_max_new_tokens: 1000
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  pad_token: <|eot_id|>

t-lite-stalin

This model is a fine-tuned version of t-bank-ai/T-lite-instruct-0.1 on the None dataset. It achieves the following results on the evaluation set:

Loss: 1.7248

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: 0.0002
train_batch_size: 48
eval_batch_size: 48
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 192
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
2.4388	0.0167	1	2.2736
1.7178	0.9874	59	1.7366
1.5568	1.9582	118	1.7031
1.4787	2.9289	177	1.7109
1.4473	3.8996	236	1.7248

Framework versions

PEFT 0.12.0
Transformers 4.44.0
Pytorch 2.3.0+cu121
Datasets 2.20.0
Tokenizers 0.19.1

alexandra-e/t-lite-stalin

t-lite-stalin

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

alexandra-e/
t-lite-stalin