File size: 3,312 Bytes
9cbece2 72e9344 9cbece2 72e9344 e545c56 ddba062 e545c56 72e9344 447424c 72e9344 447424c 72e9344 447424c 72e9344 447424c 72e9344 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 |
---
license: apache-2.0
base_model: HuggingFaceTB/cosmo-1b
tags:
- generated_from_trainer
model-index:
- name: galore-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. -->
Catastrophic forgetting test results:
Initial evaluation loss on 1k subset of HuggingFaceTB/cosmopedia-100k dataset was 1.182. 100 steps of LISA training reduced this to 1.117.
Comparison to control: cosmo-1b started out with 1.003 loss on (a different subset of) dataset, increasing to 1.024 at 100 steps.
[<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: HuggingFaceTB/cosmo-1b
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer
load_in_8bit: false
load_in_4bit: false
strict: false
datasets:
- path: Vezora/Tested-22k-Python-Alpaca
type: alpaca
dataset_prepared_path: prepared-qlora
val_set_size: 0.05
output_dir: ./galore-out
sequence_len: 2048
sample_packing: true
pad_to_sequence_len: true
adapter:
lora_model_dir:
lora_r:
lora_alpha:
lora_dropout:
lora_target_linear:
lora_fan_in_fan_out:
wandb_project: cosmo-python-galore
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:
gradient_accumulation_steps: 4
micro_batch_size: 2
num_epochs: 1
optimizer: galore_adamw_8bit
lr_scheduler: cosine
learning_rate: 0.0005
optim_target_modules:
- self_attn # for llama
- mlp
optim_args:
rank: 256
scale: 1
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
warmup_steps: 10
evals_per_epoch: 4
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
```
</details><br>
# galore-out
This model is a fine-tuned version of [HuggingFaceTB/cosmo-1b](https://huggingface.co/HuggingFaceTB/cosmo-1b) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2426
## 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: 0.0005
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_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: 1
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 0.6299 | 0.0 | 1 | 0.6469 |
| 0.4353 | 0.25 | 215 | 0.4139 |
| 0.3721 | 0.5 | 430 | 0.2957 |
| 0.3514 | 0.75 | 645 | 0.2426 |
### Framework versions
- Transformers 4.40.0.dev0
- Pytorch 2.1.2+cu118
- Datasets 2.18.0
- Tokenizers 0.15.0
|