Append 1 new entries to hyperparameters__sft (total: 1)
Browse files
README.md
CHANGED
|
@@ -1,5 +1,83 @@
|
|
| 1 |
---
|
| 2 |
dataset_info:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
- config_name: logs__llamafactory_sft
|
| 4 |
features:
|
| 5 |
- name: timestamp
|
|
@@ -107,6 +185,10 @@ dataset_info:
|
|
| 107 |
download_size: 5086
|
| 108 |
dataset_size: 225
|
| 109 |
configs:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 110 |
- config_name: logs__llamafactory_sft
|
| 111 |
data_files:
|
| 112 |
- split: train
|
|
|
|
| 1 |
---
|
| 2 |
dataset_info:
|
| 3 |
+
- config_name: hyperparameters__sft
|
| 4 |
+
features:
|
| 5 |
+
- name: model_name_or_path
|
| 6 |
+
dtype: string
|
| 7 |
+
- name: trust_remote_code
|
| 8 |
+
dtype: bool
|
| 9 |
+
- name: stage
|
| 10 |
+
dtype: string
|
| 11 |
+
- name: do_train
|
| 12 |
+
dtype: bool
|
| 13 |
+
- name: finetuning_type
|
| 14 |
+
dtype: string
|
| 15 |
+
- name: deepspeed
|
| 16 |
+
dtype: string
|
| 17 |
+
- name: dataset
|
| 18 |
+
dtype: string
|
| 19 |
+
- name: template
|
| 20 |
+
dtype: string
|
| 21 |
+
- name: cutoff_len
|
| 22 |
+
dtype: int64
|
| 23 |
+
- name: max_samples
|
| 24 |
+
dtype: int64
|
| 25 |
+
- name: overwrite_cache
|
| 26 |
+
dtype: bool
|
| 27 |
+
- name: preprocessing_num_workers
|
| 28 |
+
dtype: int64
|
| 29 |
+
- name: dataloader_num_workers
|
| 30 |
+
dtype: int64
|
| 31 |
+
- name: disable_tqdm
|
| 32 |
+
dtype: bool
|
| 33 |
+
- name: output_dir
|
| 34 |
+
dtype: string
|
| 35 |
+
- name: logging_steps
|
| 36 |
+
dtype: int64
|
| 37 |
+
- name: save_steps
|
| 38 |
+
dtype: int64
|
| 39 |
+
- name: plot_loss
|
| 40 |
+
dtype: bool
|
| 41 |
+
- name: overwrite_output_dir
|
| 42 |
+
dtype: bool
|
| 43 |
+
- name: per_device_train_batch_size
|
| 44 |
+
dtype: int64
|
| 45 |
+
- name: gradient_accumulation_steps
|
| 46 |
+
dtype: int64
|
| 47 |
+
- name: learning_rate
|
| 48 |
+
dtype: float64
|
| 49 |
+
- name: num_train_epochs
|
| 50 |
+
dtype: int64
|
| 51 |
+
- name: lr_scheduler_type
|
| 52 |
+
dtype: string
|
| 53 |
+
- name: warmup_ratio
|
| 54 |
+
dtype: float64
|
| 55 |
+
- name: weight_decay
|
| 56 |
+
dtype: float64
|
| 57 |
+
- name: adam_beta1
|
| 58 |
+
dtype: float64
|
| 59 |
+
- name: adam_beta2
|
| 60 |
+
dtype: float64
|
| 61 |
+
- name: bf16
|
| 62 |
+
dtype: bool
|
| 63 |
+
- name: ddp_timeout
|
| 64 |
+
dtype: int64
|
| 65 |
+
- name: gradient_checkpointing
|
| 66 |
+
dtype: bool
|
| 67 |
+
- name: save_only_model
|
| 68 |
+
dtype: bool
|
| 69 |
+
- name: save_strategy
|
| 70 |
+
dtype: string
|
| 71 |
+
- name: save_total_limit
|
| 72 |
+
dtype: int64
|
| 73 |
+
- name: run_name
|
| 74 |
+
dtype: string
|
| 75 |
+
splits:
|
| 76 |
+
- name: train
|
| 77 |
+
num_bytes: 500
|
| 78 |
+
num_examples: 1
|
| 79 |
+
download_size: 15735
|
| 80 |
+
dataset_size: 500
|
| 81 |
- config_name: logs__llamafactory_sft
|
| 82 |
features:
|
| 83 |
- name: timestamp
|
|
|
|
| 185 |
download_size: 5086
|
| 186 |
dataset_size: 225
|
| 187 |
configs:
|
| 188 |
+
- config_name: hyperparameters__sft
|
| 189 |
+
data_files:
|
| 190 |
+
- split: train
|
| 191 |
+
path: hyperparameters__sft/train-*
|
| 192 |
- config_name: logs__llamafactory_sft
|
| 193 |
data_files:
|
| 194 |
- split: train
|
hyperparameters__sft/train-00000-of-00001.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:657480a57211abd1d975e0d8d1df2bba78dc03d943353cfb21930231d5f5cb39
|
| 3 |
+
size 15735
|