metadata
license: gpl-3.0
library_name: peft
tags:
- generated_from_trainer
base_model: nisten/shqiponja-15b-v1
model-index:
- name: shqiponja-15
results: []
datasets:
- iamshnoo/alpaca-cleaned-albanian
- noxneural/lilium_albanicum_eng_alb
15.6b 2expert MoE
See axolotl config
axolotl version: 0.4.0
base_model: nisten/shqiponja15
model_type: AutoModelForCausalLM
tokenizer_type: LlamaTokenizer
trust_remote_code: true
load_in_8bit: false
load_in_4bit: true
strict: false
datasets:
- path: iamshnoo/alpaca-cleaned-albanian
type: alpaca
shards: 10
- path: noxneural/lilium_albanicum_eng_alb
shards: 20
type:
field_system: system
field_instruction: question
field_output: response
format: "[INST] {instruction} [/INST]"
dataset_prepared_path: last_run_prepared
val_set_size: 0.0
output_dir: ./alora-out
# - model.layers.2[7-9]+.block_sparse_moe.experts.*
# - model.layers.3[0-9]+.block_sparse_moe.experts.*
# - model.layers.2[7-9]+.b
</details><br>
# alora-out
## 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: 10
- eval_batch_size: 10
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 2
- total_train_batch_size: 80
- total_eval_batch_size: 40
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 3