|
--- |
|
license: apache-2.0 |
|
library_name: peft |
|
tags: |
|
- axolotl |
|
- generated_from_trainer |
|
base_model: unsloth/gemma-7b |
|
model-index: |
|
- name: gemma_odia_7b_unsloth |
|
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 |
|
# use google/gemma-7b if you have access |
|
base_model: unsloth/gemma-7b |
|
model_type: AutoModelForCausalLM |
|
tokenizer_type: AutoTokenizer |
|
|
|
load_in_8bit: false |
|
load_in_4bit: true |
|
strict: false |
|
|
|
# huggingface repo |
|
datasets: |
|
- path: OdiaGenAIdata/culturax-gemma-data |
|
type: completion |
|
val_set_size: 0.1 |
|
output_dir: ./gemma-odia-7b-pretrain-unsloth |
|
hub_model_id: sam2ai/gemma_odia_7b_unsloth |
|
|
|
adapter: qlora |
|
lora_r: 32 |
|
lora_alpha: 16 |
|
lora_dropout: 0.05 |
|
lora_target_linear: true |
|
|
|
sequence_len: 4096 |
|
sample_packing: true |
|
pad_to_sequence_len: true |
|
|
|
wandb_project: gemma-completion-7b-odia-unsloth |
|
wandb_entity: |
|
wandb_watch: |
|
wandb_name: |
|
wandb_log_model: |
|
|
|
|
|
gradient_accumulation_steps: 8 |
|
micro_batch_size: 2 |
|
num_epochs: 10 |
|
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: false |
|
|
|
warmup_ratio: 0.1 |
|
evals_per_epoch: 4 |
|
eval_table_size: |
|
eval_max_new_tokens: 128 |
|
saves_per_epoch: 1 |
|
debug: |
|
deepspeed: |
|
weight_decay: 0.0 |
|
fsdp: |
|
fsdp_config: |
|
special_tokens: |
|
|
|
``` |
|
|
|
</details><br> |
|
|
|
# gemma_odia_7b_unsloth |
|
|
|
This model is a fine-tuned version of [unsloth/gemma-7b](https://huggingface.co/unsloth/gemma-7b) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 2.9914 |
|
|
|
## 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: 2 |
|
- eval_batch_size: 2 |
|
- seed: 42 |
|
- distributed_type: multi-GPU |
|
- num_devices: 8 |
|
- gradient_accumulation_steps: 8 |
|
- total_train_batch_size: 128 |
|
- total_eval_batch_size: 16 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_steps: 32 |
|
- num_epochs: 10 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 39.5782 | 0.0 | 1 | 39.2579 | |
|
| 7.2511 | 0.25 | 169 | 7.0771 | |
|
| 4.2519 | 0.5 | 338 | 4.0654 | |
|
| 3.7348 | 0.75 | 507 | 3.5937 | |
|
| 3.4573 | 1.0 | 676 | 3.3126 | |
|
| 3.4299 | 1.24 | 845 | 3.2429 | |
|
| 3.4908 | 1.49 | 1014 | 3.2063 | |
|
| 3.3588 | 1.74 | 1183 | 3.1614 | |
|
| 3.3646 | 1.99 | 1352 | 3.1313 | |
|
| 3.2672 | 2.23 | 1521 | 3.0885 | |
|
| 3.2706 | 2.48 | 1690 | 3.0678 | |
|
| 3.173 | 2.73 | 1859 | 3.0410 | |
|
| 3.7319 | 2.98 | 2028 | 3.5392 | |
|
| 3.3142 | 3.22 | 2197 | 3.1610 | |
|
| 3.2931 | 3.47 | 2366 | 3.1339 | |
|
| 3.3045 | 3.72 | 2535 | 3.0710 | |
|
| 3.2423 | 3.97 | 2704 | 3.0920 | |
|
| 3.2565 | 4.2 | 2873 | 3.0311 | |
|
| 3.1167 | 4.45 | 3042 | 3.0039 | |
|
| 3.1624 | 4.71 | 3211 | 3.0108 | |
|
| 3.1697 | 4.96 | 3380 | 3.1008 | |
|
| 3.1434 | 5.19 | 3549 | 2.9915 | |
|
| 3.2301 | 5.44 | 3718 | 3.0033 | |
|
| 3.1686 | 5.69 | 3887 | 2.9893 | |
|
| 3.9959 | 5.95 | 4056 | 3.7561 | |
|
| 3.3066 | 6.18 | 4225 | 3.1076 | |
|
| 3.2567 | 6.43 | 4394 | 3.0679 | |
|
| 3.1764 | 6.68 | 4563 | 3.0459 | |
|
| 3.1848 | 6.93 | 4732 | 3.0342 | |
|
| 3.181 | 7.17 | 4901 | 3.0279 | |
|
| 3.1688 | 7.42 | 5070 | 3.0203 | |
|
| 3.1474 | 7.67 | 5239 | 3.0131 | |
|
| 3.1672 | 7.92 | 5408 | 3.0080 | |
|
| 3.1202 | 8.16 | 5577 | 3.0036 | |
|
| 3.1368 | 8.41 | 5746 | 2.9999 | |
|
| 3.1104 | 8.66 | 5915 | 2.9968 | |
|
| 3.1236 | 8.91 | 6084 | 2.9939 | |
|
| 3.1055 | 9.15 | 6253 | 2.9924 | |
|
| 3.1563 | 9.4 | 6422 | 2.9918 | |
|
| 3.1373 | 9.65 | 6591 | 2.9914 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.9.0 |
|
- Transformers 4.40.0.dev0 |
|
- Pytorch 2.4.0.dev20240326+rocm6.0 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.0 |
