Sponsored by
Agentkit
manu / dataset_1_model
Model's Last Updated: May 31 2024
text-generation
Introduction of dataset_1_model
Model Details of dataset_1_model
See axolotl config
axolotl version:
0.4.0
base_model: croissantllm/CroissantLLMBase
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizerFast
is_llama_derived_model: true
special_tokens:
bos_token: "<s>"
eos_token: "</s>"
unk_token: "<unk>"
tokens:
- "<|im_start|>"
- "<|im_end|>"
load_in_8bit: false
load_in_4bit: false
strict: false
datasets:
- path: manu/dataset_1
split: train
type: sharegpt
chat_template: "chatml"
default_system_message: ""
dataset_prepared_path: new_pii
val_set_size: 0.05
output_dir: /gpfs/workdir/fayssema/models/out_translation
sequence_len: 2048
sample_packing: false
pad_to_sequence_len: false
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: 16
num_epochs: 3
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.00003
train_on_inputs: false
group_by_length: false
bf16: auto
fp16: false
tf32: true
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: 0.05
fsdp:
fsdp_config:
gpfs/workdir/fayssema/models/out_translation
This model is a fine-tuned version of croissantllm/CroissantLLMBase on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0098
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-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.6652 | 0.0 | 1 | 2.0261 |
| 0.2986 | 0.25 | 73 | 0.0199 |
| 0.19 | 0.5 | 146 | 0.0136 |
| 0.3032 | 0.76 | 219 | 0.0158 |
| 0.1343 | 1.01 | 292 | 0.0125 |
| 0.12 | 1.26 | 365 | 0.0117 |
| 0.2266 | 1.51 | 438 | 0.0113 |
| 0.1924 | 1.77 | 511 | 0.0097 |
| 0.1448 | 2.02 | 584 | 0.0095 |
| 0.0718 | 2.27 | 657 | 0.0098 |
| 0.1184 | 2.52 | 730 | 0.0097 |
| 0.1124 | 2.77 | 803 | 0.0098 |
Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0
Runs of manu dataset_1_model on huggingface.co
7
Total runs
0
24-hour runs
0
3-day runs
0
7-day runs
0
30-day runs
More Information About dataset_1_model huggingface.co Model
Url of dataset_1_model
dataset_1_model huggingface.co Url
Provider of dataset_1_model huggingface.co
Other API from manu
Total runs:
142
Run Growth:
-4
Growth Rate:
-2.82%
Total runs:
140
Run Growth:
-51
Growth Rate:
-36.43%
Total runs:
138
Run Growth:
-4
Growth Rate:
-2.90%
Total runs:
117
Run Growth:
-20
Growth Rate:
-17.09%
Total runs:
25
Run Growth:
19
Growth Rate:
76.00%
Total runs:
24
Run Growth:
16
Growth Rate:
66.67%
Total runs:
18
Run Growth:
-10
Growth Rate:
-55.56%
Total runs:
14
Run Growth:
1
Growth Rate:
8.33%
Total runs:
7
Run Growth:
1
Growth Rate:
14.29%
Total runs:
6
Run Growth:
-8
Growth Rate:
-133.33%
Total runs:
5
Run Growth:
0
Growth Rate:
0.00%
Total runs:
5
Run Growth:
1
Growth Rate:
20.00%
Total runs:
3
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%