|
--- |
|
license: other |
|
library_name: peft |
|
tags: |
|
- axolotl |
|
- generated_from_trainer |
|
base_model: Qwen/Qwen1.5-4B |
|
model-index: |
|
- name: qwen_1.5_odia_4b |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl) |
|
<details><summary>See axolotl config</summary> |
|
|
|
axolotl version: `0.4.0` |
|
```yaml |
|
base_model: Qwen/Qwen1.5-4B |
|
model_type: AutoModelForCausalLM |
|
tokenizer_type: AutoTokenizer |
|
|
|
# is_qwen_derived_model: true |
|
trust_remote_code: true |
|
|
|
load_in_8bit: false |
|
load_in_4bit: true |
|
strict: false |
|
|
|
datasets: |
|
- path: OdiaGenAI/all_combined_odia_171k |
|
type: alpaca:chatml |
|
dataset_prepared_path: |
|
val_set_size: 0.05 |
|
output_dir: ./lora-out-qwen-4b-odia |
|
hub_model_id: sam2ai/qwen_1.5_odia_4b |
|
|
|
sequence_len: 2048 # supports up to 8192 |
|
sample_packing: false |
|
pad_to_sequence_len: |
|
|
|
adapter: qlora |
|
lora_model_dir: |
|
lora_r: 32 |
|
lora_alpha: 16 |
|
lora_dropout: 0.05 |
|
lora_target_linear: true |
|
lora_fan_in_fan_out: |
|
|
|
wandb_project: Qwen-instruct-4b-odia |
|
wandb_entity: |
|
wandb_watch: |
|
wandb_name: |
|
wandb_log_model: |
|
|
|
gradient_accumulation_steps: 2 |
|
micro_batch_size: 1 |
|
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: false |
|
early_stopping_patience: |
|
resume_from_checkpoint: |
|
local_rank: |
|
logging_steps: 1 |
|
xformers_attention: |
|
flash_attention: |
|
|
|
warmup_steps: 10 |
|
evals_per_epoch: 4 |
|
eval_table_size: |
|
eval_table_max_new_tokens: 128 |
|
saves_per_epoch: 1 |
|
debug: |
|
deepspeed: |
|
weight_decay: 0.0 |
|
fsdp: |
|
fsdp_config: |
|
special_tokens: |
|
|
|
``` |
|
|
|
</details><br> |
|
|
|
# qwen_1.5_odia_4b |
|
|
|
This model is a fine-tuned version of [Qwen/Qwen1.5-4B](https://huggingface.co/Qwen/Qwen1.5-4B) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.3421 |
|
|
|
## 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: 1 |
|
- eval_batch_size: 1 |
|
- seed: 42 |
|
- distributed_type: multi-GPU |
|
- num_devices: 8 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 16 |
|
- 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 | |
|
|:-------------:|:-----:|:-----:|:---------------:| |
|
| 0.977 | 0.0 | 1 | 1.0190 | |
|
| 0.4901 | 0.25 | 2108 | 0.4872 | |
|
| 0.3966 | 0.5 | 4216 | 0.4347 | |
|
| 0.3127 | 0.75 | 6324 | 0.4104 | |
|
| 0.3172 | 1.0 | 8432 | 0.3932 | |
|
| 0.281 | 1.25 | 10540 | 0.3778 | |
|
| 0.2845 | 1.5 | 12648 | 0.3684 | |
|
| 0.2459 | 1.75 | 14756 | 0.3616 | |
|
| 0.1641 | 2.0 | 16864 | 0.3525 | |
|
| 0.2121 | 2.25 | 18972 | 0.3506 | |
|
| 0.2564 | 2.5 | 21080 | 0.3448 | |
|
| 0.1378 | 2.75 | 23188 | 0.3426 | |
|
| 0.2002 | 3.0 | 25296 | 0.3409 | |
|
| 0.1671 | 3.25 | 27404 | 0.3439 | |
|
| 0.1464 | 3.5 | 29512 | 0.3421 | |
|
| 0.1741 | 3.75 | 31620 | 0.3421 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.8.2 |
|
- Transformers 4.37.0 |
|
- Pytorch 2.0.1+gita61a294 |
|
- Datasets 2.16.1 |
|
- Tokenizers 0.15.0 |
