ptoro's picture
Upload folder using huggingface_hub
afb6b63 verified
|
raw
history blame
3.66 kB
---
base_model: microsoft/phi-1_5
library_name: peft
license: mit
tags:
- generated_from_trainer
model-index:
- name: outputs/phi-sft-out
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: `0.4.1`
```yaml
base_model: microsoft/phi-1_5
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer
load_in_8bit: false
load_in_4bit: true
strict: false
datasets:
- path: ptoro/honkers-phi
type: alpaca
dataset_prepared_path:
val_set_size: 0.05
output_dir: ./outputs/phi-sft-out
sequence_len: 2048
sample_packing: true
pad_to_sequence_len: true
adapter: qlora
lora_model_dir:
lora_r: 64
lora_alpha: 32
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:
wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:
gradient_accumulation_steps: 1
micro_batch_size: 2
num_epochs: 4
optimizer: adamw_torch
adam_beta2: 0.95
adam_epsilon: 0.00001
max_grad_norm: 1.0
lr_scheduler: cosine
learning_rate: 0.000003
train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: true
gradient_checkpointing: true
gradient_checkpointing_kwargs:
use_reentrant: True
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
warmup_steps: 100
evals_per_epoch: 4
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.1
fsdp:
fsdp_config:
resize_token_embeddings_to_32x: true
special_tokens:
pad_token: "<|endoftext|>"
```
</details><br>
# outputs/phi-sft-out
This model is a fine-tuned version of [microsoft/phi-1_5](https://huggingface.co/microsoft/phi-1_5) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.5482
## 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: 3e-06
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 1.2333 | 0.0106 | 1 | 1.5896 |
| 1.7286 | 0.2553 | 24 | 1.5891 |
| 1.2823 | 0.5106 | 48 | 1.5875 |
| 1.3856 | 0.7660 | 72 | 1.5844 |
| 1.244 | 1.0213 | 96 | 1.5804 |
| 1.2499 | 1.2447 | 120 | 1.5753 |
| 1.1656 | 1.5 | 144 | 1.5706 |
| 1.1928 | 1.7553 | 168 | 1.5656 |
| 1.1623 | 2.0106 | 192 | 1.5608 |
| 1.2679 | 2.2340 | 216 | 1.5571 |
| 1.2845 | 2.4894 | 240 | 1.5537 |
| 1.1226 | 2.7447 | 264 | 1.5516 |
| 1.2575 | 3.0 | 288 | 1.5497 |
| 1.2465 | 3.2234 | 312 | 1.5486 |
| 1.1699 | 3.4787 | 336 | 1.5483 |
| 1.2021 | 3.7340 | 360 | 1.5482 |
### Framework versions
- PEFT 0.11.2.dev0
- Transformers 4.41.1
- Pytorch 2.1.2+cu118
- Datasets 2.19.1
- Tokenizers 0.19.1