DeepDream2045 commited on
Commit
3e5654b
1 Parent(s): a6d9588

End of training

Browse files
README.md ADDED
@@ -0,0 +1,146 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ base_model: MNC-Jihun/Mistral-7B-AO-u0.5-b2-ver0.4
4
+ tags:
5
+ - axolotl
6
+ - generated_from_trainer
7
+ model-index:
8
+ - name: 937baa9d-2d42-415a-8873-b1257eeb1985
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ [<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)
16
+ <details><summary>See axolotl config</summary>
17
+
18
+ axolotl version: `0.4.1`
19
+ ```yaml
20
+ adapter: lora
21
+ base_model: MNC-Jihun/Mistral-7B-AO-u0.5-b2-ver0.4
22
+ bf16: auto
23
+ chat_template: llama3
24
+ dataset_prepared_path: null
25
+ datasets:
26
+ - data_files:
27
+ - a1a9aab34212846d_train_data.json
28
+ ds_type: json
29
+ field: content
30
+ path: /workspace/input_data/a1a9aab34212846d_train_data.json
31
+ type: completion
32
+ debug: null
33
+ deepspeed: null
34
+ early_stopping_patience: 1
35
+ eval_max_new_tokens: 128
36
+ eval_steps: 25
37
+ eval_table_size: null
38
+ flash_attention: false
39
+ fp16: false
40
+ fsdp: null
41
+ fsdp_config: null
42
+ gradient_accumulation_steps: 16
43
+ gradient_checkpointing: true
44
+ group_by_length: true
45
+ hub_model_id: DeepDream2045/937baa9d-2d42-415a-8873-b1257eeb1985
46
+ hub_repo: null
47
+ hub_strategy: checkpoint
48
+ hub_token: null
49
+ learning_rate: 0.0001
50
+ load_in_4bit: false
51
+ load_in_8bit: false
52
+ local_rank: null
53
+ logging_steps: 1
54
+ lora_alpha: 64
55
+ lora_dropout: 0.05
56
+ lora_fan_in_fan_out: null
57
+ lora_model_dir: null
58
+ lora_r: 32
59
+ lora_target_linear: true
60
+ lr_scheduler: cosine
61
+ max_steps: 50
62
+ micro_batch_size: 2
63
+ mlflow_experiment_name: /tmp/a1a9aab34212846d_train_data.json
64
+ model_type: AutoModelForCausalLM
65
+ num_epochs: 3
66
+ optimizer: adamw_torch
67
+ output_dir: miner_id_24
68
+ pad_to_sequence_len: true
69
+ resume_from_checkpoint: null
70
+ s2_attention: null
71
+ sample_packing: false
72
+ save_steps: 25
73
+ sequence_len: 2048
74
+ strict: false
75
+ tf32: false
76
+ tokenizer_type: AutoTokenizer
77
+ train_on_inputs: false
78
+ trust_remote_code: true
79
+ val_set_size: 0.05
80
+ wandb_entity: null
81
+ wandb_mode: online
82
+ wandb_name: 937baa9d-2d42-415a-8873-b1257eeb1985
83
+ wandb_project: Gradients-On-Demand
84
+ wandb_run: your_name
85
+ wandb_runid: 937baa9d-2d42-415a-8873-b1257eeb1985
86
+ warmup_ratio: 0.05
87
+ weight_decay: 0.01
88
+ xformers_attention: true
89
+
90
+ ```
91
+
92
+ </details><br>
93
+
94
+ # 937baa9d-2d42-415a-8873-b1257eeb1985
95
+
96
+ This model is a fine-tuned version of [MNC-Jihun/Mistral-7B-AO-u0.5-b2-ver0.4](https://huggingface.co/MNC-Jihun/Mistral-7B-AO-u0.5-b2-ver0.4) on the None dataset.
97
+ It achieves the following results on the evaluation set:
98
+ - Loss: 1.7395
99
+
100
+ ## Model description
101
+
102
+ More information needed
103
+
104
+ ## Intended uses & limitations
105
+
106
+ More information needed
107
+
108
+ ## Training and evaluation data
109
+
110
+ More information needed
111
+
112
+ ## Training procedure
113
+
114
+ ### Training hyperparameters
115
+
116
+ The following hyperparameters were used during training:
117
+ - learning_rate: 0.0001
118
+ - train_batch_size: 2
119
+ - eval_batch_size: 2
120
+ - seed: 42
121
+ - distributed_type: multi-GPU
122
+ - num_devices: 4
123
+ - gradient_accumulation_steps: 16
124
+ - total_train_batch_size: 128
125
+ - total_eval_batch_size: 8
126
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
127
+ - lr_scheduler_type: cosine
128
+ - lr_scheduler_warmup_steps: 2
129
+ - training_steps: 50
130
+
131
+ ### Training results
132
+
133
+ | Training Loss | Epoch | Step | Validation Loss |
134
+ |:-------------:|:------:|:----:|:---------------:|
135
+ | 1.707 | 0.0011 | 1 | 2.0192 |
136
+ | 2.1519 | 0.0269 | 25 | 1.7509 |
137
+ | 2.1614 | 0.0538 | 50 | 1.7395 |
138
+
139
+
140
+ ### Framework versions
141
+
142
+ - PEFT 0.13.2
143
+ - Transformers 4.46.0
144
+ - Pytorch 2.4.1+cu124
145
+ - Datasets 3.0.1
146
+ - Tokenizers 0.20.1
adapter_config.json CHANGED
@@ -20,13 +20,13 @@
20
  "rank_pattern": {},
21
  "revision": null,
22
  "target_modules": [
23
- "up_proj",
24
- "q_proj",
25
- "gate_proj",
26
  "o_proj",
27
  "down_proj",
 
 
 
28
  "v_proj",
29
- "k_proj"
30
  ],
31
  "task_type": "CAUSAL_LM",
32
  "use_dora": false,
 
20
  "rank_pattern": {},
21
  "revision": null,
22
  "target_modules": [
 
 
 
23
  "o_proj",
24
  "down_proj",
25
+ "gate_proj",
26
+ "k_proj",
27
+ "up_proj",
28
  "v_proj",
29
+ "q_proj"
30
  ],
31
  "task_type": "CAUSAL_LM",
32
  "use_dora": false,
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4197433db3b3c46756d4d86e4e7cd0e492d264ae95064944875a5a8de7139dad
3
+ size 860011282
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a098e4073b8a0fba6a26374975a056a3b02d4cf25c917449c64558f8aafcf8fa
3
  size 859909312
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:641eb0d184ee62ddb72b2621ad6e78f932ea4c29eca641c702063c63c6f10f77
3
  size 859909312
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cde298347aefb5569ecda1d0ffbfc1f74fe5a8aad71a508769958efdcaf21bdf
3
  size 6776
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:987fd07a10c7dc3d0202f3fd844ca5c32e5eda76cf15397ae30934c49a3acd1e
3
  size 6776