kvasYA commited on
Commit
44ed547
1 Parent(s): e6ef5a8

End of training

Browse files
Files changed (3) hide show
  1. README.md +188 -0
  2. adapter_model.bin +3 -0
  3. adapter_model.safetensors +1 -1
README.md ADDED
@@ -0,0 +1,188 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: apache-2.0
4
+ base_model: Qwen/Qwen2.5-0.5B
5
+ tags:
6
+ - axolotl
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: ea897821-db3a-4494-836b-c96e18fa5dd9
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: Qwen/Qwen2.5-0.5B
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
+ - a758034fee445470_train_data.json
31
+ ds_type: json
32
+ format: custom
33
+ num_proc: 4
34
+ path: /workspace/input_data/a758034fee445470_train_data.json
35
+ streaming: true
36
+ type:
37
+ field_input: premise
38
+ field_instruction: hypothesis
39
+ field_output: reason
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:
47
+ lm_head: 1
48
+ model.embed_tokens: 0
49
+ model.layers.0: 0
50
+ model.layers.1: 0
51
+ model.layers.10: 1
52
+ model.layers.11: 1
53
+ model.layers.2: 0
54
+ model.layers.3: 0
55
+ model.layers.4: 0
56
+ model.layers.5: 0
57
+ model.layers.6: 1
58
+ model.layers.7: 1
59
+ model.layers.8: 1
60
+ model.layers.9: 1
61
+ model.norm: 1
62
+ do_eval: true
63
+ early_stopping_patience: 1
64
+ eval_batch_size: 1
65
+ eval_sample_packing: false
66
+ eval_steps: 25
67
+ evaluation_strategy: steps
68
+ flash_attention: false
69
+ fp16: null
70
+ fsdp: null
71
+ fsdp_config: null
72
+ gradient_accumulation_steps: 32
73
+ gradient_checkpointing: true
74
+ group_by_length: true
75
+ hub_model_id: kvasYA/ea897821-db3a-4494-836b-c96e18fa5dd9
76
+ hub_strategy: checkpoint
77
+ hub_token: null
78
+ learning_rate: 0.0001
79
+ load_in_4bit: false
80
+ load_in_8bit: false
81
+ local_rank: null
82
+ logging_steps: 1
83
+ lora_alpha: 64
84
+ lora_dropout: 0.05
85
+ lora_fan_in_fan_out: null
86
+ lora_model_dir: null
87
+ lora_r: 32
88
+ lora_target_linear: true
89
+ lora_target_modules:
90
+ - q_proj
91
+ - v_proj
92
+ lr_scheduler: cosine
93
+ max_grad_norm: 0.3
94
+ max_memory:
95
+ 0: 70GB
96
+ 1: 70GB
97
+ cpu: 96GB
98
+ max_steps: 50
99
+ micro_batch_size: 1
100
+ mixed_precision: bf16
101
+ mlflow_experiment_name: /tmp/a758034fee445470_train_data.json
102
+ model_type: AutoModelForCausalLM
103
+ num_epochs: 3
104
+ optim_args:
105
+ adam_beta1: 0.9
106
+ adam_beta2: 0.95
107
+ adam_epsilon: 1e-5
108
+ optimizer: adamw_torch
109
+ output_dir: miner_id_24
110
+ pad_to_sequence_len: true
111
+ resume_from_checkpoint: null
112
+ s2_attention: null
113
+ sample_packing: false
114
+ save_steps: 25
115
+ save_strategy: steps
116
+ sequence_len: 2048
117
+ strict: false
118
+ tf32: false
119
+ tokenizer_type: AutoTokenizer
120
+ torch_compile: false
121
+ torch_dtype: bfloat16
122
+ train_on_inputs: false
123
+ trust_remote_code: true
124
+ use_cache: false
125
+ val_set_size: 50
126
+ wandb_entity: null
127
+ wandb_mode: online
128
+ wandb_name: ea897821-db3a-4494-836b-c96e18fa5dd9
129
+ wandb_project: Public_TuningSN
130
+ wandb_runid: ea897821-db3a-4494-836b-c96e18fa5dd9
131
+ warmup_ratio: 0.05
132
+ weight_decay: 0.01
133
+ xformers_attention: null
134
+
135
+ ```
136
+
137
+ </details><br>
138
+
139
+ # ea897821-db3a-4494-836b-c96e18fa5dd9
140
+
141
+ This model is a fine-tuned version of [Qwen/Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B) on the None dataset.
142
+ It achieves the following results on the evaluation set:
143
+ - Loss: nan
144
+
145
+ ## Model description
146
+
147
+ More information needed
148
+
149
+ ## Intended uses & limitations
150
+
151
+ More information needed
152
+
153
+ ## Training and evaluation data
154
+
155
+ More information needed
156
+
157
+ ## Training procedure
158
+
159
+ ### Training hyperparameters
160
+
161
+ The following hyperparameters were used during training:
162
+ - learning_rate: 0.0001
163
+ - train_batch_size: 1
164
+ - eval_batch_size: 1
165
+ - seed: 42
166
+ - gradient_accumulation_steps: 32
167
+ - total_train_batch_size: 32
168
+ - 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
169
+ - lr_scheduler_type: cosine
170
+ - lr_scheduler_warmup_steps: 2
171
+ - training_steps: 50
172
+
173
+ ### Training results
174
+
175
+ | Training Loss | Epoch | Step | Validation Loss |
176
+ |:-------------:|:------:|:----:|:---------------:|
177
+ | 3.1199 | 0.0002 | 1 | nan |
178
+ | 2.946 | 0.0047 | 25 | nan |
179
+ | 1.6765 | 0.0095 | 50 | nan |
180
+
181
+
182
+ ### Framework versions
183
+
184
+ - PEFT 0.13.2
185
+ - Transformers 4.46.0
186
+ - Pytorch 2.5.0+cu124
187
+ - Datasets 3.0.1
188
+ - Tokenizers 0.20.1
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a870b4ebfeaa7bb5873a69530677c6707d70256fa983f493865790978bbdd8ba
3
+ size 70506570
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7c3a89f1244ba80315427e57532bb95c1755ba50904da34137a0ff443774f294
3
  size 70430032
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:405c497fbd0979cfd44e56d4927eccc1821e1771804b5877af2d328ae77a9636
3
  size 70430032