cilooor commited on
Commit
f0ee115
·
verified ·
1 Parent(s): 09198ec

End of training

Browse files
Files changed (2) hide show
  1. README.md +171 -0
  2. adapter_model.bin +3 -0
README.md ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: apache-2.0
4
+ base_model: Maykeye/TinyLLama-v0
5
+ tags:
6
+ - axolotl
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: 39171b5e-8255-4afe-b007-e37b2055ba48
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ [<img src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/axolotl-ai-cloud/axolotl)
17
+ <details><summary>See axolotl config</summary>
18
+
19
+ axolotl version: `0.4.1`
20
+ ```yaml
21
+ adapter: lora
22
+ base_model: Maykeye/TinyLLama-v0
23
+ bf16: true
24
+ chat_template: llama3
25
+ data_processes: 16
26
+ dataset_prepared_path: null
27
+ datasets:
28
+ - data_files:
29
+ - 99a7259d7db7ed53_train_data.json
30
+ ds_type: json
31
+ format: custom
32
+ path: /workspace/input_data/99a7259d7db7ed53_train_data.json
33
+ type:
34
+ field_input: input
35
+ field_instruction: instruction
36
+ field_output: output
37
+ format: '{instruction} {input}'
38
+ no_input_format: '{instruction}'
39
+ system_format: '{system}'
40
+ system_prompt: ''
41
+ debug: null
42
+ deepspeed: null
43
+ device_map: auto
44
+ early_stopping_patience: 5
45
+ eval_batch_size: 4
46
+ eval_max_new_tokens: 128
47
+ eval_steps: 50
48
+ eval_table_size: null
49
+ evals_per_epoch: null
50
+ flash_attention: false
51
+ fp16: false
52
+ fsdp: null
53
+ fsdp_config: null
54
+ gradient_accumulation_steps: 4
55
+ gradient_checkpointing: true
56
+ group_by_length: true
57
+ hub_model_id: cilooor/39171b5e-8255-4afe-b007-e37b2055ba48
58
+ hub_repo: null
59
+ hub_strategy: checkpoint
60
+ hub_token: null
61
+ learning_rate: 2.0e-05
62
+ load_in_4bit: false
63
+ load_in_8bit: false
64
+ local_rank: null
65
+ logging_steps: 1
66
+ lora_alpha: 128
67
+ lora_dropout: 0.05
68
+ lora_fan_in_fan_out: null
69
+ lora_model_dir: null
70
+ lora_r: 64
71
+ lora_target_linear: true
72
+ lr_scheduler: cosine
73
+ lr_scheduler_warmup_steps: 10
74
+ max_grad_norm: 0.5
75
+ max_memory:
76
+ 0: 75GB
77
+ max_steps: 200
78
+ micro_batch_size: 4
79
+ mlflow_experiment_name: /tmp/99a7259d7db7ed53_train_data.json
80
+ model_type: AutoModelForCausalLM
81
+ num_epochs: 3
82
+ optim_args:
83
+ adam_beta1: 0.9
84
+ adam_beta2: 0.999
85
+ adam_epsilon: 1e-8
86
+ optimizer: adamw_bnb_8bit
87
+ output_dir: miner_id_24
88
+ pad_to_sequence_len: true
89
+ resume_from_checkpoint: null
90
+ s2_attention: null
91
+ sample_packing: false
92
+ save_steps: 250
93
+ saves_per_epoch: null
94
+ seed: 42
95
+ sequence_len: 1024
96
+ special_tokens:
97
+ pad_token: </s>
98
+ strict: false
99
+ tf32: true
100
+ tokenizer_type: AutoTokenizer
101
+ total_train_batch_size: 16
102
+ train_batch_size: 8
103
+ train_on_inputs: false
104
+ trust_remote_code: true
105
+ val_set_size: 0.05
106
+ wandb_entity: null
107
+ wandb_mode: online
108
+ wandb_name: 4901f9ea-986a-429f-a444-29d313ed458c
109
+ wandb_project: Gradients-On-Demand
110
+ wandb_run: your_name
111
+ wandb_runid: 4901f9ea-986a-429f-a444-29d313ed458c
112
+ warmup_steps: 10
113
+ weight_decay: 0.01
114
+ xformers_attention: null
115
+
116
+ ```
117
+
118
+ </details><br>
119
+
120
+ # 39171b5e-8255-4afe-b007-e37b2055ba48
121
+
122
+ This model is a fine-tuned version of [Maykeye/TinyLLama-v0](https://huggingface.co/Maykeye/TinyLLama-v0) on the None dataset.
123
+ It achieves the following results on the evaluation set:
124
+ - Loss: 7.9445
125
+
126
+ ## Model description
127
+
128
+ More information needed
129
+
130
+ ## Intended uses & limitations
131
+
132
+ More information needed
133
+
134
+ ## Training and evaluation data
135
+
136
+ More information needed
137
+
138
+ ## Training procedure
139
+
140
+ ### Training hyperparameters
141
+
142
+ The following hyperparameters were used during training:
143
+ - learning_rate: 2e-05
144
+ - train_batch_size: 4
145
+ - eval_batch_size: 4
146
+ - seed: 42
147
+ - gradient_accumulation_steps: 4
148
+ - total_train_batch_size: 16
149
+ - optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=adam_beta1=0.9,adam_beta2=0.999,adam_epsilon=1e-8
150
+ - lr_scheduler_type: cosine
151
+ - lr_scheduler_warmup_steps: 10
152
+ - training_steps: 200
153
+
154
+ ### Training results
155
+
156
+ | Training Loss | Epoch | Step | Validation Loss |
157
+ |:-------------:|:------:|:----:|:---------------:|
158
+ | 12.5588 | 0.0001 | 1 | 11.7775 |
159
+ | 8.9314 | 0.0031 | 50 | 9.2843 |
160
+ | 7.7455 | 0.0063 | 100 | 8.2949 |
161
+ | 7.4296 | 0.0094 | 150 | 7.9863 |
162
+ | 7.4188 | 0.0125 | 200 | 7.9445 |
163
+
164
+
165
+ ### Framework versions
166
+
167
+ - PEFT 0.13.2
168
+ - Transformers 4.46.0
169
+ - Pytorch 2.5.0+cu124
170
+ - Datasets 3.0.1
171
+ - Tokenizers 0.20.1
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34af278a712646dcdfbae05430b2dfe8964279d87fa9c1c55a005999fc99ee4c
3
+ size 3054730