dsakerkwq commited on
Commit
3bd7c8e
1 Parent(s): 1239bcb

End of training

Browse files
Files changed (3) hide show
  1. README.md +176 -0
  2. adapter_model.bin +3 -0
  3. adapter_model.safetensors +1 -1
README.md ADDED
@@ -0,0 +1,176 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: apache-2.0
4
+ base_model: llamafactory/tiny-random-Llama-3
5
+ tags:
6
+ - axolotl
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: 962a6f10-da4a-4fa3-859b-54e1a8a79f0b
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: llamafactory/tiny-random-Llama-3
23
+ bf16: auto
24
+ chat_template: llama3
25
+ cosine_min_lr_ratio: 0.1
26
+ data_processes: 4
27
+ dataset_prepared_path: null
28
+ datasets:
29
+ - data_files:
30
+ - 279d709c0e410472_train_data.json
31
+ ds_type: json
32
+ format: custom
33
+ num_proc: 4
34
+ path: /workspace/input_data/279d709c0e410472_train_data.json
35
+ streaming: true
36
+ type:
37
+ field_input: candidates
38
+ field_instruction: problem
39
+ field_output: answer
40
+ format: '{instruction} {input}'
41
+ no_input_format: '{instruction}'
42
+ system_format: '{system}'
43
+ system_prompt: ''
44
+ debug: null
45
+ deepspeed: null
46
+ device_map: balanced
47
+ do_eval: true
48
+ early_stopping_patience: 1
49
+ eval_batch_size: 1
50
+ eval_sample_packing: false
51
+ eval_steps: 25
52
+ evaluation_strategy: steps
53
+ flash_attention: false
54
+ fp16: null
55
+ fsdp: null
56
+ fsdp_config: null
57
+ gradient_accumulation_steps: 16
58
+ gradient_checkpointing: true
59
+ group_by_length: true
60
+ hub_model_id: dsakerkwq/962a6f10-da4a-4fa3-859b-54e1a8a79f0b
61
+ hub_strategy: checkpoint
62
+ hub_token: null
63
+ learning_rate: 0.0001
64
+ load_in_4bit: false
65
+ load_in_8bit: false
66
+ local_rank: null
67
+ logging_steps: 1
68
+ lora_alpha: 64
69
+ lora_dropout: 0.05
70
+ lora_fan_in_fan_out: null
71
+ lora_model_dir: null
72
+ lora_r: 32
73
+ lora_target_linear: true
74
+ lora_target_modules:
75
+ - q_proj
76
+ - v_proj
77
+ lr_scheduler: cosine
78
+ max_grad_norm: 1.0
79
+ max_memory:
80
+ 0: 75GB
81
+ 1: 75GB
82
+ 2: 75GB
83
+ 3: 75GB
84
+ max_steps: 50
85
+ micro_batch_size: 2
86
+ mixed_precision: bf16
87
+ mlflow_experiment_name: /tmp/279d709c0e410472_train_data.json
88
+ model_type: AutoModelForCausalLM
89
+ num_epochs: 3
90
+ optim_args:
91
+ adam_beta1: 0.9
92
+ adam_beta2: 0.95
93
+ adam_epsilon: 1e-5
94
+ optimizer: adamw_torch
95
+ output_dir: miner_id_24
96
+ pad_to_sequence_len: true
97
+ resume_from_checkpoint: null
98
+ s2_attention: null
99
+ sample_packing: false
100
+ save_steps: 25
101
+ save_strategy: steps
102
+ sequence_len: 2048
103
+ special_tokens:
104
+ pad_token: <|eot_id|>
105
+ strict: false
106
+ tf32: false
107
+ tokenizer_type: AutoTokenizer
108
+ torch_compile: false
109
+ train_on_inputs: false
110
+ trust_remote_code: true
111
+ val_set_size: 50
112
+ wandb_entity: null
113
+ wandb_mode: online
114
+ wandb_name: 962a6f10-da4a-4fa3-859b-54e1a8a79f0b
115
+ wandb_project: Public_TuningSN
116
+ wandb_runid: 962a6f10-da4a-4fa3-859b-54e1a8a79f0b
117
+ warmup_ratio: 0.04
118
+ weight_decay: 0.01
119
+ xformers_attention: null
120
+
121
+ ```
122
+
123
+ </details><br>
124
+
125
+ # 962a6f10-da4a-4fa3-859b-54e1a8a79f0b
126
+
127
+ This model is a fine-tuned version of [llamafactory/tiny-random-Llama-3](https://huggingface.co/llamafactory/tiny-random-Llama-3) on the None dataset.
128
+ It achieves the following results on the evaluation set:
129
+ - Loss: 11.7333
130
+
131
+ ## Model description
132
+
133
+ More information needed
134
+
135
+ ## Intended uses & limitations
136
+
137
+ More information needed
138
+
139
+ ## Training and evaluation data
140
+
141
+ More information needed
142
+
143
+ ## Training procedure
144
+
145
+ ### Training hyperparameters
146
+
147
+ The following hyperparameters were used during training:
148
+ - learning_rate: 0.0001
149
+ - train_batch_size: 2
150
+ - eval_batch_size: 1
151
+ - seed: 42
152
+ - distributed_type: multi-GPU
153
+ - num_devices: 4
154
+ - gradient_accumulation_steps: 16
155
+ - total_train_batch_size: 128
156
+ - total_eval_batch_size: 4
157
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=adam_beta1=0.9,adam_beta2=0.95,adam_epsilon=1e-5
158
+ - lr_scheduler_type: cosine
159
+ - lr_scheduler_warmup_steps: 2
160
+ - training_steps: 35
161
+
162
+ ### Training results
163
+
164
+ | Training Loss | Epoch | Step | Validation Loss |
165
+ |:-------------:|:------:|:----:|:---------------:|
166
+ | 11.7667 | 0.0870 | 1 | 11.7628 |
167
+ | 11.7239 | 2.1739 | 25 | 11.7333 |
168
+
169
+
170
+ ### Framework versions
171
+
172
+ - PEFT 0.13.2
173
+ - Transformers 4.46.0
174
+ - Pytorch 2.5.0+cu124
175
+ - Datasets 3.0.1
176
+ - Tokenizers 0.20.1
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:340d10f5acfec64b48c71288b4940d1bc5f8d963460241b29972571c576afef6
3
+ size 104322
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8f097a99aa7268b76fbccb38abaec5107cbc630a03bce54bbaf4a58b1c9b54ae
3
  size 97728
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cad25fe9759dee23ee5e221a886b5ff047e7dd6b4ac5b26cb3c0bee2b1ce757
3
  size 97728