Hugging Face's logo

Alphatao
/
645a0255-97b1-4563-888a-322df2ff7346

PEFT
Safetensors
llama
axolotl
Generated from Trainer
Model card Files
xet
 Community
Alphatao commited on
Commit
bdce640
·
verified ·
1 Parent(s): 361d89c

End of training

Browse files
Files changed (3) hide show
  1. README.md +163 -0
  2. adapter_model.bin +3 -0
  3. adapter_model.safetensors +1 -1
README.md ADDED
@@ -0,0 +1,163 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: apache-2.0
4
+ base_model: unsloth/tinyllama-chat
5
+ tags:
6
+ - axolotl
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: 645a0255-97b1-4563-888a-322df2ff7346
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/tinyllama-chat
23
+ bf16: true
24
+ chat_template: llama3
25
+ dataset_prepared_path: null
26
+ datasets:
27
+ - data_files:
28
+ - a5037f04fb3737b9_train_data.json
29
+ ds_type: json
30
+ format: custom
31
+ path: /workspace/input_data/a5037f04fb3737b9_train_data.json
32
+ type:
33
+ field_input: content
34
+ field_instruction: instruction
35
+ field_output: message
36
+ format: '{instruction} {input}'
37
+ no_input_format: '{instruction}'
38
+ system_format: '{system}'
39
+ system_prompt: ''
40
+ debug: null
41
+ device_map:
42
+ ? ''
43
+ : 0,1,2,3,4,5,6,7
44
+ early_stopping_patience: 2
45
+ eval_max_new_tokens: 128
46
+ eval_steps: 100
47
+ eval_table_size: null
48
+ flash_attention: true
49
+ gradient_accumulation_steps: 8
50
+ gradient_checkpointing: true
51
+ group_by_length: false
52
+ hub_model_id: Alphatao/645a0255-97b1-4563-888a-322df2ff7346
53
+ hub_repo: null
54
+ hub_strategy: null
55
+ hub_token: null
56
+ learning_rate: 0.0002
57
+ load_best_model_at_end: true
58
+ load_in_4bit: false
59
+ load_in_8bit: false
60
+ local_rank: null
61
+ logging_steps: 1
62
+ lora_alpha: 128
63
+ lora_dropout: 0.1
64
+ lora_fan_in_fan_out: null
65
+ lora_model_dir: null
66
+ lora_r: 64
67
+ lora_target_linear: true
68
+ lora_target_modules:
69
+ - q_proj
70
+ - k_proj
71
+ - v_proj
72
+ - o_proj
73
+ lr_scheduler: cosine
74
+ max_grad_norm: 1.0
75
+ max_steps: 2520
76
+ micro_batch_size: 4
77
+ mlflow_experiment_name: /tmp/a5037f04fb3737b9_train_data.json
78
+ model_type: AutoModelForCausalLM
79
+ num_epochs: 2
80
+ optimizer: adamw_bnb_8bit
81
+ output_dir: miner_id_24
82
+ pad_to_sequence_len: true
83
+ resume_from_checkpoint: null
84
+ s2_attention: null
85
+ sample_packing: false
86
+ save_steps: 100
87
+ sequence_len: 2048
88
+ strict: false
89
+ tf32: true
90
+ tokenizer_type: AutoTokenizer
91
+ train_on_inputs: false
92
+ trust_remote_code: true
93
+ val_set_size: 0.05
94
+ wandb_entity: null
95
+ wandb_mode: online
96
+ wandb_name: 210f0dff-565a-4f98-88c9-fa4dc67322f6
97
+ wandb_project: Gradients-On-Demand
98
+ wandb_run: your_name
99
+ wandb_runid: 210f0dff-565a-4f98-88c9-fa4dc67322f6
100
+ warmup_steps: 10
101
+ weight_decay: 0.0
102
+ xformers_attention: null
103
+
104
+ ```
105
+
106
+ </details><br>
107
+
108
+ # 645a0255-97b1-4563-888a-322df2ff7346
109
+
110
+ This model is a fine-tuned version of [unsloth/tinyllama-chat](https://huggingface.co/unsloth/tinyllama-chat) on the None dataset.
111
+ It achieves the following results on the evaluation set:
112
+ - Loss: 0.6140
113
+
114
+ ## Model description
115
+
116
+ More information needed
117
+
118
+ ## Intended uses & limitations
119
+
120
+ More information needed
121
+
122
+ ## Training and evaluation data
123
+
124
+ More information needed
125
+
126
+ ## Training procedure
127
+
128
+ ### Training hyperparameters
129
+
130
+ The following hyperparameters were used during training:
131
+ - learning_rate: 0.0002
132
+ - train_batch_size: 4
133
+ - eval_batch_size: 4
134
+ - seed: 42
135
+ - gradient_accumulation_steps: 8
136
+ - total_train_batch_size: 32
137
+ - optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
138
+ - lr_scheduler_type: cosine
139
+ - lr_scheduler_warmup_steps: 10
140
+ - training_steps: 1295
141
+
142
+ ### Training results
143
+
144
+ | Training Loss | Epoch | Step | Validation Loss |
145
+ |:-------------:|:------:|:----:|:---------------:|
146
+ | 2.6724 | 0.0015 | 1 | 2.2758 |
147
+ | 1.4673 | 0.1545 | 100 | 0.6642 |
148
+ | 1.1353 | 0.3090 | 200 | 0.6505 |
149
+ | 1.1465 | 0.4635 | 300 | 0.6303 |
150
+ | 0.2895 | 0.6180 | 400 | 0.6214 |
151
+ | 1.1199 | 0.7725 | 500 | 0.6155 |
152
+ | 1.0039 | 0.9270 | 600 | 0.6042 |
153
+ | 0.2547 | 1.0815 | 700 | 0.6155 |
154
+ | 0.4809 | 1.2360 | 800 | 0.6140 |
155
+
156
+
157
+ ### Framework versions
158
+
159
+ - PEFT 0.13.2
160
+ - Transformers 4.46.0
161
+ - Pytorch 2.5.0+cu124
162
+ - Datasets 3.0.1
163
+ - Tokenizers 0.20.1
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4e9e9cfeaaaebef8d09500dc7aa443e1f156b8623c5c1be54bd87365f4100be6
3
+ size 201962138
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4db6aeff8981d28b47d148f4245546d2de4ab56075eb4ed814d8cfdbc57313c5
3
  size 201892112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e83af09b5d8fad20bbc2d3d3faf2228501518e5c5fd4700afa860e30c8f7277
3
  size 201892112