dsakerkwq commited on
Commit
7847e61
1 Parent(s): 2ee8292

End of training

Browse files
Files changed (2) hide show
  1. README.md +192 -0
  2. adapter_model.bin +3 -0
README.md ADDED
@@ -0,0 +1,192 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: llama3.1
4
+ base_model: unsloth/Llama-3.1-Storm-8B
5
+ tags:
6
+ - axolotl
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: 81ee6aa8-1abc-4905-8446-40b88b66ce39
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: unsloth/Llama-3.1-Storm-8B
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
+ - e8ca0ac66aa11e96_train_data.json
31
+ ds_type: json
32
+ format: custom
33
+ num_proc: 4
34
+ path: /workspace/input_data/e8ca0ac66aa11e96_train_data.json
35
+ streaming: true
36
+ type:
37
+ field_instruction: Hausa
38
+ field_output: English
39
+ format: '{instruction}'
40
+ no_input_format: '{instruction}'
41
+ system_format: '{system}'
42
+ system_prompt: ''
43
+ debug: null
44
+ deepspeed: null
45
+ device_map:
46
+ lm_head: 3
47
+ model.embed_tokens: 0
48
+ model.layers.0: 0
49
+ model.layers.1: 0
50
+ model.layers.10: 3
51
+ model.layers.11: 3
52
+ model.layers.2: 0
53
+ model.layers.3: 1
54
+ model.layers.4: 1
55
+ model.layers.5: 1
56
+ model.layers.6: 2
57
+ model.layers.7: 2
58
+ model.layers.8: 2
59
+ model.layers.9: 3
60
+ model.norm: 3
61
+ do_eval: true
62
+ early_stopping_patience: 1
63
+ eval_batch_size: 1
64
+ eval_sample_packing: false
65
+ eval_steps: 25
66
+ evaluation_strategy: steps
67
+ flash_attention: false
68
+ fp16: null
69
+ fsdp: null
70
+ fsdp_config: null
71
+ gradient_accumulation_steps: 32
72
+ gradient_checkpointing: true
73
+ group_by_length: true
74
+ hub_model_id: dsakerkwq/81ee6aa8-1abc-4905-8446-40b88b66ce39
75
+ hub_strategy: checkpoint
76
+ hub_token: null
77
+ learning_rate: 0.0001
78
+ load_in_4bit: false
79
+ load_in_8bit: false
80
+ local_rank: null
81
+ logging_steps: 1
82
+ lora_alpha: 64
83
+ lora_dropout: 0.05
84
+ lora_fan_in_fan_out: null
85
+ lora_model_dir: null
86
+ lora_r: 32
87
+ lora_target_linear: true
88
+ lora_target_modules:
89
+ - q_proj
90
+ - v_proj
91
+ lr_scheduler: cosine
92
+ max_grad_norm: 0.3
93
+ max_memory:
94
+ 0: 60GB
95
+ 1: 70GB
96
+ 2: 70GB
97
+ 3: 70GB
98
+ cpu: 96GB
99
+ max_steps: 50
100
+ micro_batch_size: 1
101
+ mixed_precision: bf16
102
+ mlflow_experiment_name: /tmp/e8ca0ac66aa11e96_train_data.json
103
+ model_type: AutoModelForCausalLM
104
+ num_epochs: 3
105
+ optim_args:
106
+ adam_beta1: 0.9
107
+ adam_beta2: 0.95
108
+ adam_epsilon: 1e-5
109
+ optimizer: adamw_torch
110
+ output_dir: miner_id_24
111
+ pad_to_sequence_len: true
112
+ resume_from_checkpoint: null
113
+ s2_attention: null
114
+ sample_packing: false
115
+ save_steps: 25
116
+ save_strategy: steps
117
+ sequence_len: 2048
118
+ strict: false
119
+ tf32: false
120
+ tokenizer_type: AutoTokenizer
121
+ torch_compile: false
122
+ torch_dtype: bfloat16
123
+ train_on_inputs: false
124
+ trust_remote_code: true
125
+ use_cache: false
126
+ val_set_size: 50
127
+ wandb_entity: null
128
+ wandb_mode: online
129
+ wandb_name: 81ee6aa8-1abc-4905-8446-40b88b66ce39
130
+ wandb_project: Public_TuningSN
131
+ wandb_runid: 81ee6aa8-1abc-4905-8446-40b88b66ce39
132
+ warmup_ratio: 0.05
133
+ weight_decay: 0.01
134
+ xformers_attention: null
135
+
136
+ ```
137
+
138
+ </details><br>
139
+
140
+ # 81ee6aa8-1abc-4905-8446-40b88b66ce39
141
+
142
+ This model is a fine-tuned version of [unsloth/Llama-3.1-Storm-8B](https://huggingface.co/unsloth/Llama-3.1-Storm-8B) on the None dataset.
143
+ It achieves the following results on the evaluation set:
144
+ - Loss: nan
145
+
146
+ ## Model description
147
+
148
+ More information needed
149
+
150
+ ## Intended uses & limitations
151
+
152
+ More information needed
153
+
154
+ ## Training and evaluation data
155
+
156
+ More information needed
157
+
158
+ ## Training procedure
159
+
160
+ ### Training hyperparameters
161
+
162
+ The following hyperparameters were used during training:
163
+ - learning_rate: 0.0001
164
+ - train_batch_size: 1
165
+ - eval_batch_size: 1
166
+ - seed: 42
167
+ - distributed_type: multi-GPU
168
+ - num_devices: 4
169
+ - gradient_accumulation_steps: 32
170
+ - total_train_batch_size: 128
171
+ - total_eval_batch_size: 4
172
+ - 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
173
+ - lr_scheduler_type: cosine
174
+ - lr_scheduler_warmup_steps: 2
175
+ - training_steps: 50
176
+
177
+ ### Training results
178
+
179
+ | Training Loss | Epoch | Step | Validation Loss |
180
+ |:-------------:|:------:|:----:|:---------------:|
181
+ | 0.0 | 0.0091 | 1 | nan |
182
+ | 0.0 | 0.2283 | 25 | nan |
183
+ | 0.0 | 0.4566 | 50 | nan |
184
+
185
+
186
+ ### Framework versions
187
+
188
+ - PEFT 0.13.2
189
+ - Transformers 4.46.0
190
+ - Pytorch 2.5.0+cu124
191
+ - Datasets 3.0.1
192
+ - Tokenizers 0.20.1
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e1e22a7c24473ae2ce03415b098f515aeed520202a76c93116fabb734bf0b47
3
+ size 335706186