Sponsored by
APIMaster
YoungPanda / up
Model's Last Updated: February 15 2024
text-generation
Introduction of up
Model Details of up
See axolotl config
axolotl version:
0.4.0
base_model: upstage/SOLAR-10.7B-v1.0
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer
is_llama_derived_model: true
trust_remote_code: true
load_in_8bit: false
load_in_4bit: false
strict: false
datasets:
- path: Drewskidang/share
type: sharegpt
- path: Drewskidang/shareinstruct
type: sharegpt
dataset_prepared_path: last_run_prepared
val_set_size: 0.05
output_dir: ./out
sequence_len: 4096
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:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:
gradient_accumulation_steps: 2
micro_batch_size: 8
num_epochs: 5
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: true
flash_attn_cross_entropy: false
flash_attn_rms_norm: true
flash_attn_fuse_qkv: false
flash_attn_fuse_mlp: true
warmup_steps: 100
evals_per_epoch: 4
eval_table_size:
saves_per_epoch: 1
debug:
deepspeed: #deepspeed_configs/zero2.json # multi-gpu only
weight_decay: 0.1
fsdp:
fsdp_config:
special_tokens:
out
This model is a fine-tuned version of upstage/SOLAR-10.7B-v1.0 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.5945
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: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.3027 | 0.1 | 1 | 1.3229 |
| 1.2121 | 0.29 | 3 | 1.1924 |
| 1.0946 | 0.57 | 6 | 1.1220 |
| 1.0204 | 0.86 | 9 | 1.0590 |
| 0.8295 | 1.1 | 12 | 1.0489 |
| 0.8185 | 1.38 | 15 | 1.0353 |
| 0.7114 | 1.67 | 18 | 1.0193 |
| 0.7053 | 1.95 | 21 | 1.0263 |
| 0.3707 | 2.19 | 24 | 1.1041 |
| 0.339 | 2.48 | 27 | 1.1846 |
| 0.3449 | 2.76 | 30 | 1.1659 |
| 0.3301 | 3.05 | 33 | 1.1935 |
| 0.148 | 3.29 | 36 | 1.4335 |
| 0.1682 | 3.57 | 39 | 1.3037 |
| 0.1855 | 3.86 | 42 | 1.3231 |
| 0.0857 | 4.1 | 45 | 1.4003 |
| 0.113 | 4.38 | 48 | 1.5945 |
Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.1.1+cu121
- Datasets 2.17.0
- Tokenizers 0.15.0
Runs of YoungPanda up on huggingface.co
8
Total runs
0
24-hour runs
0
3-day runs
0
7-day runs
0
30-day runs
More Information About up huggingface.co Model
Url of up
up huggingface.co Url
Provider of up huggingface.co
Other API from YoungPanda
Total runs:
6
Run Growth:
0
Growth Rate:
0.00%
Total runs:
6
Run Growth:
0
Growth Rate:
0.00%
Total runs:
2
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%