Romain-XV commited on
Commit
499879a
·
verified ·
1 Parent(s): 9f1d068

End of training

Browse files
Files changed (3) hide show
  1. README.md +173 -0
  2. adapter_model.bin +3 -0
  3. adapter_model.safetensors +1 -1
README.md ADDED
@@ -0,0 +1,173 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: apache-2.0
4
+ base_model: NousResearch/Yarn-Mistral-7b-64k
5
+ tags:
6
+ - axolotl
7
+ - generated_from_trainer
8
+ model-index:
9
+ - name: 440533d0-4ce2-46c1-8a0e-f65c18162efd
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: NousResearch/Yarn-Mistral-7b-64k
23
+ bf16: true
24
+ chat_template: llama3
25
+ dataset_prepared_path: null
26
+ datasets:
27
+ - data_files:
28
+ - 37c8687d026e9fd5_train_data.json
29
+ ds_type: json
30
+ format: custom
31
+ path: /workspace/input_data/37c8687d026e9fd5_train_data.json
32
+ type:
33
+ field_instruction: question
34
+ field_output: process
35
+ format: '{instruction}'
36
+ no_input_format: '{instruction}'
37
+ system_format: '{system}'
38
+ system_prompt: ''
39
+ debug: null
40
+ deepspeed: null
41
+ early_stopping_patience: 4
42
+ eval_max_new_tokens: 128
43
+ eval_steps: 150
44
+ eval_table_size: null
45
+ flash_attention: false
46
+ fp16: null
47
+ fsdp: null
48
+ fsdp_config: null
49
+ gradient_accumulation_steps: 4
50
+ gradient_checkpointing: true
51
+ group_by_length: false
52
+ hub_model_id: Romain-XV/440533d0-4ce2-46c1-8a0e-f65c18162efd
53
+ hub_repo: null
54
+ hub_strategy: checkpoint
55
+ hub_token: null
56
+ learning_rate: 0.0002
57
+ load_best_model_at_end: true
58
+ load_in_4bit: true
59
+ load_in_8bit: false
60
+ local_rank: null
61
+ logging_steps: 1
62
+ lora_alpha: 128
63
+ lora_dropout: 0.3
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
+ lr_scheduler: cosine
73
+ max_grad_norm: 1.0
74
+ max_steps: 2639
75
+ micro_batch_size: 2
76
+ mlflow_experiment_name: /tmp/37c8687d026e9fd5_train_data.json
77
+ model_type: AutoModelForCausalLM
78
+ num_epochs: 3
79
+ optimizer: adamw_bnb_8bit
80
+ output_dir: miner_id_24
81
+ pad_to_sequence_len: true
82
+ resume_from_checkpoint: null
83
+ s2_attention: null
84
+ sample_packing: false
85
+ save_steps: 150
86
+ sequence_len: 2048
87
+ special_tokens:
88
+ pad_token: </s>
89
+ strict: false
90
+ tf32: true
91
+ tokenizer_type: AutoTokenizer
92
+ train_on_inputs: false
93
+ trust_remote_code: true
94
+ val_set_size: 0.03300482530545966
95
+ wandb_entity: null
96
+ wandb_mode: online
97
+ wandb_name: a93fe1fb-147c-44b3-88ca-b0335c410502
98
+ wandb_project: Gradients-On-Demand
99
+ wandb_run: your_name
100
+ wandb_runid: a93fe1fb-147c-44b3-88ca-b0335c410502
101
+ warmup_steps: 10
102
+ weight_decay: 0.0
103
+ xformers_attention: null
104
+
105
+ ```
106
+
107
+ </details><br>
108
+
109
+ # 440533d0-4ce2-46c1-8a0e-f65c18162efd
110
+
111
+ This model is a fine-tuned version of [NousResearch/Yarn-Mistral-7b-64k](https://huggingface.co/NousResearch/Yarn-Mistral-7b-64k) on the None dataset.
112
+ It achieves the following results on the evaluation set:
113
+ - Loss: 0.0961
114
+
115
+ ## Model description
116
+
117
+ More information needed
118
+
119
+ ## Intended uses & limitations
120
+
121
+ More information needed
122
+
123
+ ## Training and evaluation data
124
+
125
+ More information needed
126
+
127
+ ## Training procedure
128
+
129
+ ### Training hyperparameters
130
+
131
+ The following hyperparameters were used during training:
132
+ - learning_rate: 0.0002
133
+ - train_batch_size: 2
134
+ - eval_batch_size: 2
135
+ - seed: 42
136
+ - gradient_accumulation_steps: 4
137
+ - total_train_batch_size: 8
138
+ - optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
139
+ - lr_scheduler_type: cosine
140
+ - lr_scheduler_warmup_steps: 10
141
+ - training_steps: 2639
142
+
143
+ ### Training results
144
+
145
+ | Training Loss | Epoch | Step | Validation Loss |
146
+ |:-------------:|:------:|:----:|:---------------:|
147
+ | 2.6189 | 0.0001 | 1 | 0.7597 |
148
+ | 1.8377 | 0.0082 | 150 | 0.3992 |
149
+ | 1.1386 | 0.0164 | 300 | 0.3287 |
150
+ | 0.8881 | 0.0246 | 450 | 0.2975 |
151
+ | 1.0054 | 0.0328 | 600 | 0.2704 |
152
+ | 1.0347 | 0.0410 | 750 | 0.2467 |
153
+ | 0.7466 | 0.0491 | 900 | 0.2219 |
154
+ | 0.7175 | 0.0573 | 1050 | 0.1994 |
155
+ | 0.8256 | 0.0655 | 1200 | 0.1825 |
156
+ | 0.731 | 0.0737 | 1350 | 0.1682 |
157
+ | 0.5574 | 0.0819 | 1500 | 0.1521 |
158
+ | 0.4409 | 0.0901 | 1650 | 0.1386 |
159
+ | 0.4958 | 0.0983 | 1800 | 0.1268 |
160
+ | 0.4718 | 0.1065 | 1950 | 0.1164 |
161
+ | 0.3277 | 0.1147 | 2100 | 0.1075 |
162
+ | 0.4109 | 0.1229 | 2250 | 0.1008 |
163
+ | 0.3703 | 0.1311 | 2400 | 0.0974 |
164
+ | 0.4028 | 0.1393 | 2550 | 0.0961 |
165
+
166
+
167
+ ### Framework versions
168
+
169
+ - PEFT 0.13.2
170
+ - Transformers 4.46.0
171
+ - Pytorch 2.5.0+cu124
172
+ - Datasets 3.0.1
173
+ - Tokenizers 0.20.1
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a5b5e135123d2f5e1043e10dd91d452897789b271ec6c66777bd5d136221a6e5
3
+ size 671250634
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6326305b02c5345bbc9a927c7664228a65ff343e0847e6f540371df214898245
3
  size 671149168
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:49c1378ef4736f2f7bbb0d8e6d86b1193a3f82cecb1bcad549ca99671c48a0b1
3
  size 671149168